AEHR is the rare small cap that has top line growth coupled with bottom line strength. We recently discussed what comprises a defensible portfolio in tech in our recent Q3 Kickoff webinar. AEHR fits the criteria we outlined in the webinar, and is one of our largest holdings because of how many boxes it ticks.
It’s both the consistent revenue growth andand the bottom-line growth that causes this stock to defy the odds. The FY2024 guide is for 50% growth on the top line and 90% growth on the bottom line. In addition to this fiscal year 2023 cash flow grew over 500% from $1.51 million in FY2022 to $10M in FY2023.
In our pre-earnings write-up we had stated: “Aehr is within $1m of last year’s entire GAAP earnings by the end of Q3, so with the forecasted revenues for fiscal Q4 conservative net profit on revenue, Aehr will clearly grow net profit and GAAP earnings at a higher percentage than revenue growth.”
The revenue growth can be seasonal, which is why we had said “All Eyes on FY2024 Guide” in our last earnings review. The fiscal year guide in July tends to be a baseline as it’s likely the company receives more orders in the next few months and fills these in FY2024.
I believe management is doing the responsible thing – guiding to what they know today. Because the company has a small amount of revenue, a minor miss can be substantial. To avoid this, they lean “conservative.” This was discussed in the earnings call (see below).
Financials:
FY2023 and Q4 revenue were both in line at $65 million and $22.3 million, respectively. This represents FY revenue growth of 28% and quarterly revenue growth of 9.9%. EPS also came in as expected at $0.59 and $0.23, on an adjusted basis.
The FY2024 guide represents growth of 50% for revenue of $100 million. The two analysts covering the stock had a consensus of $102.5 million for growth of 58%. The FY2024 guide on the bottom line is for GAAP net income of $28 million and at least 90% in earnings growth.
Margins were flat to slightly contracted. 51.5% on gross margin was flat from a year ago. Operating margin of 25.2% was slightly lower compared to 28.7% a year ago. Net margin of 27.3% was also lower compared to 28.6% last year.
Regarding the margins, management pointed toward R&D as a primary reason the margins were down, plus product mix, material and transportation costs. At the current FY2024 guide, management is implying flat margins YoY and that’s a positive given the GAAP margins are very strong (i.e., opposite of “growth at any cost”, this is growth while maintaining a strong bottom line). G&A will increase this year as AEHR is a $1 billion market cap and auditing costs are expected to increase.
The company has $47.9 million in cash, up from $31.5 million a year ago and up from $4.6M two years ago. Operating cash flow of $5.8 million in Q4 was up considerably from (-$0.77) million a year ago. For AEHR, the exact cash flow figures are available when the company files their 10-Q/10-K. Since it's the end of a fiscal year, AEHR will be filing a 10-K.
Last quarter, AEHR had reported $33.3 million, the highest bookings in company history. The fiscal year-to-date was $72.5 million, exceeding full prior fiscal year of $62.2 million. The current effective backlog is $40 million with $15 million added in the first six weeks of the fiscal year, which started June 1st.
Earnings Call:
The main questions to focus on from the earnings call was regarding number of customers and customer concentration, hints toward if the guide was purposely low, plus product optionality with gallium nitride, higher power voltages and silicon photonics.
Number of Customers and Customer Concentration:
AEHR’s top risk remains its high customer concentration. When asked about this on the forum, I’ve stated it’s different from semiconductor companies compared to cloud software or fintech, for example. For a semiconductor company like Aehr, the supply chain limits the number of customers they have to probably a dozen or so for its total addressable market. This is different than a software company that will have thousands of customers at scale.
Regardless, it’s a risk and one we’ve covered many times. In the call, it was discussed that the primary customer (which is generally known to be ON Semiconductor) represents 79% of revenue. When we first covered AEHR, ON was about 100% of revenue.
In FY2023, the customer concentration improved to three customers representing: 79%/10%/10%. In other words, AEHR had two 10% customers. In FY2024, the company is expected to have “three or four” 10% customers.
Overall, the discussions around AEHR’s customer growth are quite bullish and likely contributed to the positive price action. For example, this was stated:
“With the addition of this latest new customer, we've significantly expanded our customer base by adding a total of four new silicon carbide customers this year. Each of these new customers is already ramping or plans to ramp our products into high volume production using our multi-wafer test and burn-in systems.
We believe this customer who serves several significant markets that include the electric vehicle industry as well as other industrial applications will purchase a large number of our FOX-XP systems to meet their publicly announced significant increase in plant capacity and revenue growth over the next several years and through the end of the decade and longer.”
Questions on FY2024 Guidance:
It was discussed (by an analyst) that the guide should be particularly easy to hit given the customer mix. The analyst felt the $40 million in backlog was especially where the guide/current information is too low.
“And specifically, if we look at this past year and your largest customer being 52 million based on the 79% and their targets of growing 300% over the next few years. I would assume that creates a solid base for your business. So, as you look at already having 8 million to 10 million booked on that, number two, you're really talking about a 30 million incremental number to hit that minimum threshold and you've already got three customers.”
Management’s response was “I felt like you were going to end it with why we're such sandbaggers but anyhow” and then went on to state that timing orders is too difficult for their business in order to guide aggressively: “It's interesting even with current customers candidly their ability to forecast is all over the map. And so, I think we've taken a conservative stance here. But it provides us with confidence to be able to hit that number. And we don't need any miracles to happen, if you will.
Total Addressable Market:
AEHR is a company where a few different TAMs are thrown around in the earnings calls. All of them are sizable, and perhaps the highest TAMs of any company we cover relative to AEHR’s size, and the products being in a unique niche with virtually no competitors.
We’ve discussed the TAM many times but here is what was most recently stated:
“William Blair forecast total demand for silicon carbide wafer is just for electric vehicles, which include EV, inverters onboard and offboard chargers to grow from 220,000 wafers in 2022 to over 4.5 million six-inch equivalent wafers in 2030, a greater than 45% compound annual growth rate and over 20 times larger in 2030 than in 2022.
In addition, William Blair expects demand for industrial applications, trains, energy conversion and RF amplifiers of silicon carbide to drive another 2.8 million wafers in 2030. This expands our silicon carbide test and burn-in market even more.”
Combined, the market for AEHR could be as large as 7.3 million wafers, up from 220,000. This is 33.2X growth. Notably, this assumes AEHR takes the entire market, which is not likely to happen even with a superior product. The product is superior because wafer level testing is up to 9 times faster than competitors as customers can test and burn-in/stabilize nine 300mm wafers at the same time compared to one wafer with competitors at 3.5 kilowatts of power per wafer.
Product Optionality:
There were discussions about AEHR’s next two major markets, silicon photonics and gallium nitride. For background on these two new markets, which would extend AEHR’s total addressable market beyond silicon carbide, please read this analysis here.
Per the earnings call, AEHR has officially received their first silicon photonics order whereas gallium nitride is in the more nascent phase of “customer inquiries.”
This system can test new high power density devices that can be used in new optical I/O or heterogeneous integrated packages. This customer is one of the world's largest semiconductor manufacturers and we expect to receive orders for additional production systems as they have increased production of these devices.”
Conclusion:
With AEHR, we broke a few of our portfolio rules. First, we have held a small cap at a high allocation. Secondly, we are holding it beyond the 10% allocation limit most portfolios (including the I/O Fund) adhere to for risk management purposes. Third, we bought close to earnings. It’s not ideal to break this many portfolio rules unless the stock is special. Clearly, we think AEHR is special. To reiterate, it’s special not only for its top line growth but primarily for its bottom-line growth, total addressable market and product optionality.
This article was originally published on Forbes on Forbes Forbes on Jun 29, 2023, 08:52 pm EDT
The market is like the weather, it changes often. The market’s fickle nature is partly why stocks often lead to losses for retail investors. By going “all-in’ or “all-out,” individual investors can often be overexposed to the sudden changes the market brings. This is especially true when the market has treated investors well as it creates a sense of security – or worse, complacency.
Our site is unique in that we provide active management. This has helped us outperform across four audit periods. Our stance is the weather will always change – from good times to bad times, and from bad times to good times. Therefore, we are not over exposed in either direction.
An example of this is Nvidia, our largest position. As the allocation grew well beyond 10%, we took gains. Even after taking gains, the company is currently at a 17% allocation. In May and June, I stated that our firm was not buying Nvidia right now. Well, similar to the weather, this has changed. My firm bought a small tranche of Nvidia yesterday for $410. This tranche will come with a stop, meaning if the stock sells off, we will close this 2% tranche while hedging the 15% original position.
Sign up for I/O Fund's free newsletter with gains of up to 221% – Click hereClick here
Our Current Nvidia Trade:
With the cash we raised throughout 2022, NVDA was the primary target of deploying some of this cash once our analysis signaled a bottom was in place. The below is a real-time trade notification we sent to our members on the October 13th.
Source: I/O FUND
The above alert was 1 of 9 alerts we sent out from 2021 – 2022 to buy NVDA below $200. However, since February of 2023, we have been systematically taking gains at key levels based on technical and macro warnings. Even with logging sizable wins while raising cash, it in the top position in 2023.
In our pre-earnings buy-plan for NVDA, we stated that “It is our belief that NVDA is setting up for a sizable pullback, which we believe will open the door for better long-term entries.” Though we do believe that lower levels will manifest in time,the recent earnings report moved forward expectations regarding AI, which is showing up in the price action. We have been discussing that Nvidia will be an AI leader for years with an allocation to match, yet predicting the exact day and month the market would finally price in this thesis is impossible to predict (and timing to this level is not necessary when holding a large longer-term position)
Regarding price, we work in probabilities, and when the market changes, so do we. The key to NVDA today is the large gap from their earnings report. This gap is either a breakaway gap, or an exhaustion gap. If it is a breakaway gap, which is represented by our red count below, then it is the halfway point in this push higher. On the other hand, if price breaks below $340, likely on some type of “event,” then the gap is an exhaustion gap, and will mark a larger top. This is represented by our blue count.
Every Thursday at 4:30 pm Eastern, the I/O Fund team holds a webinar for premium members to discuss how to navigate the broad market, as well as various stock entries and exits. We offer trade alerts plus an automated hedging signal. The I/O Fund team is one of the only audited portfolios available to individual investors. Learn more here.Learn more here.
Source: I/O Fund
The $405 – $395 region will likely continue to act as strong support for the red count. This is where we added back in anticipation for a ~38% push higher. Our stop for this move will be a break below the $340 critical support region, which is ~14% lower than our entry.
Unlike many, we do not believe AI is a bubble, nor do we think the valuations in some of these names is stretched, as many believe. What does concern us regarding possible “events” are: 1) geo-political tensions forcing a ban of selling NVDA’s chips to China – which Beth spoke about in May with Bloomberg Asia; 2) the inevitable recession that will likely start to be priced into equities in Q4/Q1, but could get pushed forward due to an unforeseen event.
Because of these risks, we are buying with an exit plan for any new entries. It is our belief, based on the economic data, that a recession is a more likely than not for the US economy. However, based on current projections on timing, we could see a continued push in AI leadership through year-end. This is what we are further positioning our portfolio for, with the realization that we could top out sooner than anticipated.
The I/O Fund has been beating the drum about AI for 5 years. Now that it is here, we are targeting choice mid-cap to mega-cap names in the coming pullback. Once this exuberance runs its course, and the market gives up on AI, we will be buying the dip for this once-in-a-lifetime tech trend that is just starting. Join us the week following the holiday, Thursday, 7/13, at 4:30 EST where we will go over the specific AI stocks we are targeting. We will provide the macro backdrop, along with entry prices.about AI for 5 years. Now that it is here, we are targeting choice mid-cap to mega-cap names in the coming pullback. Once this exuberance runs its course, and the market gives up on AI, we will be buying the dip for this once-in-a-lifetime tech trend that is just starting. Join us the week following the holiday, Thursday, 7/13, at 4:30 EST where we will go over the specific AI stocks we are targeting. We will provide the macro backdrop, along with entry prices.
Last year, the I/O Fund reinitiated a position into AEHR and published a deep dive entitled: “AEHR Analysis: The Silicon Carbide Revolution.” This is a must-read for anyone new to our site. Currently, the position is up over 100% YTD and is second only to our Nvidia position, up 170% YTD. Notably, Aehr was also the primary winner for the portfolio in 2022, based a series in entries beginning in August.
ON Semiconductor is Aehr’s number one customer. We wanted to take a look at ON to see if it has a place in our portfolio. This analysis also helps to provide a 360-degree view of the silicon carbide market.
As a refresher on silicon carbide, below are some notes from an analysis we published in August.
Recap of why Silicon Carbide, also known as “SiC,” is disrupting Silicon-insulated gate bipolar transistors (Si-IGBTs):
The result in switching to Silicon Carbide (SiC) is that charging is quicker and the range of miles for electric vehicles increases with SiC. Si-IGBTs are inefficient, oversized and have trouble achieving pure sine wave voltage requirements whereas Silicon Carbide can withstand and manage high voltages. This is a good fit for electric vehicles which have high-voltage batteries.
When you replace silicon with silicon carbide, the breakdown strength increases 10X and can operate at higher temperatures and provide higher current density. SiC devices offer 3X more thermal conductivity and allow for faster heat dissipation. As silicon devices become smaller, it’s more difficult to extract the heat from the electrical conversion process.
By replacing silicon with silicon carbide in MOSFETs, the low switching losses and higher switching frequencies are retained. Due to the durability of silicon carbide, MOSFETs are now also able to handle higher voltage at lower heat. Notably, silicon carbide combines silicon with carbon and is the third-hardest substance in the world. The durability of silicon carbide is also ideal for the various conditions electric vehicles must operate in and the design is also more compact.
By withstanding higher temperatures combined with lower switching losses and lower thermal resistance, silicon carbide (SiC) can handle more power while using less energy. SiC reduces the power consumption and reduces the size of power supply systems that require high-voltage conversion, which makes SiC especially compatible with electric vehicle (EV) on-board chargers and solar photovoltaic power systems.
Tesla was the first to adopt silicon carbide for the 2018 Model 3 by working with ST Microelectronics to add SiC MOSFETs to an inverter design. The result was a more compact, lighter inverter at 4.8Kg compared to Si IGBT inverters that weigh 2-3X more (8kg to 12kg). Published in the New York Times: “Tesla made this fantastic move,” said Claire Troadec, an analyst at Yole Développement, a high-tech research and consulting firm in France, referring to the company’s switch to silicon carbide. “What they did in a year and a half was really amazing.
Pictured below, the main roadblock to SiC MOSFETs adoption is cost yet this was largely solved for as the second-gen Tesla’s SiC inverters, which analysts believe are now comparable to Si-IGBTs. Not only is the SiC inverter on par in cost but is known to be one of the best on the market at 97% efficiency, resulting in more range. This was accomplished without increasing battery capacity.
Aehr provides the testing equipment necessary to ensure the reliability of silicon carbide devices. The stakes are high should an electric vehicle or solar panels fail in the field, considering not only the costs involved with these products ($50,000+ for EVs, $10,000+ for solar systems) but it also protects the reputation of a particular brand in a competitive environment. AEHR’s testing equipment provides the necessary step of quality assurance. The testing equipment is also used to increase battery life before going to market.
Aehr has a unique technology called FOX-XP test systems, which are used for wafer level burn-in testing of silicon carbide and silicon photonics. As stated, ON Semi is the number one customer for Aehr’s test systems and wafer paks.
To learn more about Aehr, please reference the following analysis:
ON Semiconductor provides the silicon carbide MOSFETs (which are semiconductors that switch or supply voltages) for electric vehicle components for EV drive trains, such as traction inverters, DC/DC inverters, on-board chargers, fast chargers and energy storage applications.
ON is Aehr’s number one customer because they must test silicon carbide semiconductors for failures before supplying silicon carbide-based powertrains and charging systems for key Electric Vehicle OEMs, such as Tesla, Volkswagen, Mercedes Benz, Nio and BMW.
EliteSiC is the portfolio of SiC products that ON provides to popular EV OEMs to increase electric power train efficiency and extend EV range. Of the estimated $8.13 billion in revenue that On Semi is expected to report this year FY2023, $1 billion is from silicon carbide components. This is up 5X from 2022. In Q1, the SiC revenue doubled from Q4.
Here is where the other $7 billion in revenue comes from:
Intelligent sensors used in automotive ADAS systems, which are abundant in Level 1+ autonomous vehicle systems.
Note: ON Semi has the #1 global position in image sensors with a 46% market share in Autos and 68% market share in ADAS. This particular segment accounts for 69% of total revenue, up from 64% in the year ago quarter.
Smart factory automation and robotics
Energy infrastructure, such as EV charging stations
Retail applications, such as for surveillance or doorbell cameras
Cloud and 5G infrastructure
The company has been ambitious in vertically integrating the manufacturing process to reduce dependencies. For example, ON Semi acquired GT Advanced Technology in 2021, a SiC manufacturer, which eliminates the need to rely on a partner to source SiC. The company also acquired a 300mm fab from Global Foundries to strengthen its imaging sensor business. Both facilities are located in the US and also benefit from tax credits from the CHIPS act.
What’s Driving Growth:
The key markets underpinning the growth in the Intelligent Power and Sensing businesses are the Automotive and Industrial sectors which comprise over 75% of total revenue. Of which Auto is 50% and Industrial is 28%. Importantly, the similarity in customer requirements – increased power and efficiency – provide engineering and product development synergies between the two businesses.
In the chart above, ON expects to grow 3x the market and forecasts an average sales growth of 10-12% from 2022 through 2027.
Intelligent Sensing
Within Intelligent Sensing, this Advanced Driver Assistance Systems (ADAS) schematic from Spiceworks demonstrates the impact car electronification has had on car safety with the radar, light detection and ranging (LIDAR), cameras and embedded ultrasound technology.
In Image Sensors, ON has the #1 global position in autos and industrials. It has a 46% market share in Autos and 68% in ADAS. Below are examples of cameras embedded on ON chips used in automotive and industrial applications.
The quality (i.e. megapixels) and integration of these sensing technologies is one of ON’s key advantages. For example, it can operate at lower temperature operation, has reduced cooling requirements and weighs less.
In addition, the range and ability to function in low light environments are also differentiating factors. This is how ON has described the market drivers: “We introduced our new Hyperlux Family of image sensors to support the transition to eight megapixel devices where ASPs (average sales prices) can be up to 2.5 times that of one or two megapixel image sensors.”
ON was the first to market with 8 megapixel automotive-grade sensors that provide both high detection range and a wide field of view, delivering consistent performance across all temperature and lighting conditions. Based on this industry-leading dynamic range, dark noise performance and LED flicker mitigation feature of its sensor, ON has been winning new designs. Last year ON displaced the large incumbent at local Japanese automotive OEMs.
In addition, one of the primary drivers of increasing number of cameras per car has been the efforts by traditional OEMs to match the ADAS and related safety features offered by the new EV models. For example, ON secured a design win for a digital mirror that incorporates four cameras for the rear and outside views. This mirror overcomes obstructions caused by passengers, headrests and other objects and provides integrated rearview and side view for blind spot monitoring.
In industrial end markets, growth is driven by industrial and warehouse automation applications where ON has 27% market share. ON’s image sensors are used for scanning in industrial and warehouse applications. This growth is driven by expansion and increased automation of warehouse by global e-commerce leaders. Customers include Amazon on warehouses and Schneider, Siemens, ABB and Honeywell in factories.
ON is the only image sensor supplier with internal and external capabilities across every manufacturing stage of the supply chain for automotive and industrial sensors.
Intelligent Power
Within Intelligent Power, ON’s key product is its Silicon Carbide technology. In Silicon Carbide and Silicon power, ON has #2 Global position with 9% market share.
Beyond traction inverters for electric vehicles, the SiC solution can be applied across several different end markets where AC/DC or DC/DC power conversion is needed including high speed trains, photovoltaic converters, medical equipment, motor drives, electric power transmission and solid state transformers.
ON’s EV customers require SiC technology that meets two basic requirements, greater range with less weight. Take for example, ON’s EV clients such as Tesla and VW. They require SiC technology that will provide greater efficiency, increased voltage and greater power density, which reduces charging times and increases driving ranges between charges.
ON has a broad portfolio of product solutions to meet EV maker needs.
ON’s EliteSiC and VE-Trac module solutions provide the foundation for traction inverter design, making automotive products more powerful and reliable. Main traction inverters are the heart of electric vehicles and provide incredible amounts of torque and acceleration. The responsiveness of the inverter and the electric motor it controls correlate directly to the “feel” of the vehicle and consumer satisfaction.
Power levels from 40 kW to 250+ kW are common, and these systems require extremely robust IGBT and silicon carbide (SiC) components. Scalability, enhanced thermal performance, and the low packaging inductance allow ON’s traction inverters to achieve peak efficiencies, state-of-art power density, and swift response times.
EV makers seek increased range and reliability but with less size weight in their SiC requirements. ON is currently on M3 silicon carbide technology and is developing M4 and M5. The key is that each new generation is able to reduce the size of the cell but increase the amount of current that goes through it.
The packaging around the Die is also very important. The packaging has to effectively allow heat to be extracted from the Die, through the packaging and into the fluid and be able to operate at higher temperatures. For example, ON’s gel-based Case Modules are a good choice for applications up to 150 degrees Celsius. However, EV based SiC based modules require packaging such as Transfer Molded that can function at 200 degrees Celsius. By reducing thermal resistance, this enables more power, efficiency and better range at a lower cost.
Importantly, it weighs less. For example, a gel-based Case Module weighs 5 lbs., while a metal Dual Side Transfer Molded weighs 50% less but with higher power performance at higher temperatures. This reduces the weight of the vehicle. ON recently developed polymer-based package which weighs 90% less than the metal based. Packaging solutions can be “mixed and matched” depending on the need.
Longer-Term Supply Arrangements (LTSA)
As part of ON’s strategy, it seeks to enter into longer-term supply arrangements (LTSA) with strategic end-customers. Typically, they last from 4 to 5 years. LTSA’s provide a level of visibility on pricing and volume through the duration of the arrangement.
Coupled with the value-added products that ON is providing, these arrangements typically lead to a more stable pricing environment. ON provides a quarterly update on the level of LTSAs which provides a good proxy of future growth.
ON ended q123 with $17.6B in LTSA’s, of which about $4.5b is silicon carbide related. To provide some context, ON’s 2022 revenue was $8.3bn.
Importantly, the LTSA’s provide a roadmap as to when and how much to expand capacity. Rather, than building capacity with the hope that they can fill it. ON can now build capacity to fulfill its LTSAs. This reduces the margin volatility inherent in the “build it and hope they will come” model. In addition, it has helped ON to provide accurate profitability guidance to wall street analysts.
Per management on the earnings call:
“LTSAs also help reduce our exposure to the volatility in the consumer and computing markets. Volkswagen as an example, signed a three-year agreement for more than 100 current production devices giving them the required supply chain sustainability with a major semiconductor partner, our committed revenue through LTSAs increased again in Q1 by $1 billion.”
Key Earnings Driver: Increasing ON Content Per Product
The key to ON’s earnings and continued growth will be the increase of its semiconductor content in automotive and industrial applications in its key growth markets as well as in units shipped.
Below is a diagram showing the typical ON content in a higher end German EV. Silicon Carbide, Silicon Power components and Power ICs are supplied by the Intelligent Power division. While Sensor Interfaces and Image Sensors are supplied by Intelligent Sensing.
Currently, the value of ON content for electric vehicles ranges from $350-$1,800, depending on the type of car. As EV penetration increases, ON will continue to benefit both from the increase in content per car and the number of EV cars made. Importantly, the former helps reduce the cyclicality of the latter (i.e. Auto Sales SAAR).
Per the earnings call: “We also lead the market in automotive ultrasonic sensors with more than 20 sensors in one of the latest EV models from a leading European OEMs.”
An important factor to point out is that the cost of ON’s content relative to its client’s overall costs for its product is fairly small. The costs of ON’s EV components are only about 2.7% of the total COGS for an EV manufacturer while the Advanced Safety components are a bit less than that, compared to $40,000 total COGS for an EV maker. Yet, the importance to the operation and quality of the car is much more valuable than that. This should provide ON a healthy level of pricing power in its LTSAs.
Recent Acquisitions
GTAT acquisition
In November 2021, ON completed its acquisition of Hudson, New Hampshire based GT Advanced Technologies, a manufacturer of Silicon Carbide (SiC).
With GTAT, ON is now vertically integrated and canprovide end-to-end power solutions from SiC crystal growth to fully integrated intelligent power modules. ON is the only integrated device manufacturer that grows its own silicon carbide and silicon boules.
This vertical integration enables ON to maximize production efficiency between the 3 facilities. Since the acquisition, substrate production is 10x higher, die up 12x, internal packing 4x, yields 1.7x and new product 3x.
East Fishkill Fab – Global Foundries
ON is the only image sensor supplier with internal and external capabilities across every manufacturing stage of the supply chain for automotive and industrial sensors.
In February 2023, ON finalized its acquisition of Global Foundries’ 300 mm fab in East Fishkill, NY (EFK) which will enhance its competitive position by enabling the conversion of ON’s wafer processes from 200mm to 300mm. The fab increases the processing capabilities required for image sensor production.
Management/CEO:
Hassane El-Khoury has been CEO since December 2020. Prior to that, he was the President and CEO of Cypress Semiconductor before its sale to Infineon. Since joining, he has increased ON’s focus toward higher growth/higher margin markets, price maximization across its product offerings and increased vertical integration.
CEO El-Khoury spearheaded the transition and divestment from sub-scale factories overseas into a lighter internal fabrication model with the goal of reducing gross margin volatility while at the same time making strategic acquisitions, such as the GT Advanced Technology acquisition and the East Fishkill Fab acquisition from Global Foundries – actions all with an eye toward increasing gross margins while reducing margin volatility inherent to the sector.
So far, CEO El-Khoury has a track record of saying what he will do and doing what he says. In doing so, he has built credibility with the investment community and Wall Street. Since December 2020, ON has outperformed the SOX by 140% and 250% outperformance compared to Wolfspeed through May 2023.
ON has characterized 2023 as a transition year, mainly due to integration of its acquisitions and divestment of its non-core assets. We see 50%+ valuation potential when this transition is completed by 2h23 (see more on valuation below).
Financials:
ON Semi had a streak of strong revenue growth in the 20% range that has recently flatlined to the single-digit negative growth expected for the remaining quarters this year. The expectation is that ON Semi returns to growth in FY2024 at 7.5% revenue growth.
EPS returns to growth in FY2024, as well, at +11.74% for $5.43 EPS. This compares to (-8.77%) EPS decline in FY2023 to $4.86.
Margins:
Gross Margin has been in the 49% range yet has currently dipped to 46.8% in the most recent quarter. The guidance is for 45.4% to 47.4%. This is a much better margin than competitor Wolfspeed in the 30% range.
The operating margin was 28.8% last quarter and net margin was 23.6%. This is also much better than competitor Wolfspeed, a company with double digit negative profit margins.
ON has guided for gross margins to be lower in 2023 vs 2022. ON has characterized 2023 as a transition year on the path toward a resumption of higher margins.
The main headwinds are:
Continued elimination of 10-15% of sales (~$700m) from non-core/lower margins businesses first announced at the end of 2021 to be concluded by 2023
Lower fab utilization to manage inventories at its EFK facility
Start-up costs related to its SiC Hudson and facilities which will weigh on margins
Ultimately, silicon carbide components are helping margins yet the East Fishkill fab is weighing on margins. Per Management, the headwinds are temporary: “Based on our current outlook, we are confident we can realign the cost structure of the fab and drive efficiencies to recover by early 2024.”
Cash Flow:
Last year, ON had an operating cash flow margin of 31.60% and a free cash flow margin of 19.60%. This margin has been fairly consistent until this quarter when it dropped to 20.90% for op cash flow and 4.50% for free cash flow margin. There is $2.7 billion in cash on the balance sheet and $3.46 billion in debt.
The free cash flow is softer from “lost” net income from higher inventory, or in other words, the elimination of sales.
Per management, regarding inventory increasing QoQ: “Inventory increased by $198.1 million sequentially and days of inventory increased by 23 days to 159 days. This includes approximately 43 days of bridge inventory to support fab transitions and the impending silicon carbide ramp. We continue to proactively manage distribution inventory, decreasing inventory in the channel by $79 million sequentially and at historically low levels with weeks of inventory at 7 weeks, compared to 7.3 weeks in Q4.”
Management also stated they are directing “a significant portion” of capex toward silicon carbide and the 300mm capabilities of the East Fishkill fab.
Key Metrics:
Last quarter, the automotive business grew 38% year-over-year to $986 million yet was flat QoQ. The YoY growth is in line with the year ago quarter, also at 38% growth YoY. The company guided to automotive being “positive” quarter-over-quarter in the upcoming quarter.
Industrial grew 2% in the March quarter to $556 million in revenue while the “Other” category declined (-39.3%) from $687 million to $417 million.
Here is a breakdown of the revenue segments:
Valuation:
ON Semi is trading higher than its historic average on the top line. The PS Ratio is 4.6 compared to 5-year average of 2.56.
The PE Ratio of 20.89 is lower than the 5-year average of 30, yet notably this valuation has been range bound in the 16-20 range since early 2022.
ON’s valuation is not demanding relative to the market and semiconductor sector indicating that the market has not yet looked past 2023. Once the market gets indications that the 2023 year of transition is almost over, it can begin to focus on 2025 eps estimates and the 50%+ valuation potential.
Per the table below, currently, consensus is modeling $4.86 (2023), $5.43 (2024), and $6.39 (2025). At the moment, the earnings potential is 31% from 2023 to 2025. The 2023 estimates were negative y/y at the beginning of the year but they have increased post the better than expected q123 and the analyst day. So, the positive eps revisions (albeit the absolute eps is still down y/y) are a positive factor.
Our goal for an entry would be to see 50% upside in price from $86 to ~ $130. The rationale is that once ON gets through the 2023 transition, analysts will turn toward the 2025 valuation based on $6.39 2025 EPS where currently it trades 14x. Putting 20x multiple on that gets you to around $130 (before discounting).
At the investor day, ON gave targets out to 2027 with ON guiding for operating margin expansion from 35% to 40% by 2027.
Earnings Call:
Question on East Fishkill dragging on Margins:
“On the gross margin side of things. It seems like there were quite a few moving parts, especially the East Fishkill and on the silicon carbide side, but the net of it all seems be right in line with your plan. Can you just talk a little bit about those moving parts? East Fishkill is more expensive, but silicon carbide is ahead of plan. Does that still net out to the same trajectory through the rest of the year? Just walk us through those puts and takes and maybe the utilization side as part of that as well, please?”
Thad Trent, CFO:
“Yes. So the utilization dropped in the quarter from about 74% to 71%. We expect, kind of, what we're seeing right now is utilization to stay in that range plus or minus for the remainder of the year. Obviously, if there's a second-half recovery, we can ramp up quickly. You nailed it on the rest of it silicon carbide performed better-than-expected, EFK cost as I said is coming in significantly higher than we expected. You can think about these as being, kind of, orders of magnitude more dilutive than what we expected.
The good news is we are absorbing that. As I said, we're finding additional opportunities to improve gross margin across the company and we're able to absorb that. We believe by the time we get into 2024, we've got the cost structure of EFK back in line to where we would expect it to be. So we're really confident in the margin outlook for this year, I don't think anything changes. I think if we look at Street consensus for gross margin for 2023, even with these headwinds, we think we can execute to that — those expectations.”
Question on When the Company Will Return to Growth:
“Yes, thank you and congratulations as well on great results in a tough environment. Just a quick question on the automotive market, you mentioned Hassane, a modest inventory digestion in the end market and then you're also kind of reducing distribution inventory. Can you talk a little bit about the overall demand picture for automotive. I know it's hard to kind of separate the significant ramp that you're seeing in electric vehicles and in turn silicon carbide. But just curious if there's a softness in the demand market, if there's a shift away from high-end to mid-range, any kind of color on the automotive market will be appreciated?”
Hassane El-Khoury:
Yes, look. We don't see a big disconnect into demand. It was like I said, it was a momentary thing where we use this opportunity to kind of reposition the inventory that we have externally and we'll get back to growth in the second quarter and through the year giving us an increase in our automotive revenue year-over-year.
Question on Lead Times at the Company:
Question:
Anyway, can you just give us a little update and some color on the shortages and the lead time situation? I guess for Hassane, our shortage is pretty much exclusively in the automotive business or are they elsewhere? And then there any point in time this year where you think the shortages will go away?
Hassane El-Khoury
Yes, look, so the — for me, I always refer to shortages as technologies, because they're across all markets where we provide them. High voltage silicon is of course constrained technology for us. We ramped capacity, yet the demand is much higher than even our increased capacity. And for that business, for example, it goes into automotive and it goes into industrial specifically in our alternative energy. And as Thad said, that's ramping very nicely this year after a very stellar ‘22 ramp that we talked about last year. So that is technology that is constrained. We have some intelligent power technologies that are constrained. Think about it as mixed signal analog where demand in automotive and demand in industrial both have been increasing ahead of the capacity we've added.
Past Performance:
At the end of 2021, ON provided medium term financial targets. In particular a 45% Gross Margin target by 2025, exiting 10-15% (~$700m) of non-core revenue by 2022-2023 and optimizing manufacturing to increase profitability.
By the end of 2022, ON had reached and exceeded those 2021 profitability targets ahead of schedule. In particular, gross margins reached 49%. In doing so, ON has been establishing a track record of saying what they will do and then doing better than that which has helped management build credibility with the market as reflected in the stock’s performance.
From 2020 to 2022, ON’s revenue growth was 26% CAGR and Gross Margins improved by 1,310 bps and free cash flow increased significantly.
“In Q1 alone, these results allowed us to nearly double our Q4 revenue and more than half of our 2022 full-year revenue. We are on track to grow our revenue to $1 billion in 2023 and that's approximately 5 times the revenue of 2022 setting ourselves up for leadership in the silicon carbide market with the majority of the substrate sourced internally.
Demand for electric vehicles, ADAS and energy infrastructure remained healthy amid a broad-based macroeconomic slowdown. While our automotive revenue increased 38% year-over-year, it was flat quarter-over-quarter. We are still supply constrained across several automotive technologies, while in some other technologies, we are cautiously monitoring inventory digestion.”
Investor’s Day and 2027 Targets
Following the May Q123 earnings call, ON held an investor day later in the month.
ON provided new financial updates through to 2027. A gross margin target of 53% based on 1) product mix, 2) estimated growth rate of 10-12% from 2022 through 2027 in its Automotive and Industrial markets 3) contribution from SiC and 4) manufacturing efficiencies and cost savings.
Per an analyst note, the 10-12% annually through 2027 is three-times the semiconductor industry growth.
ON also guided for operating margins to go from 34.7% to 40% in the same time period.
How does this impact the short and medium term profitability?
Taking into consideration these moving pieces, we can look at the impact on Wall Street consensus estimates. In 2022, ON had sales of $8.33b and normalized EPS of $5.33. In 2021, sales were $6.74b and normalized EPS of $2.95.
For 2023, analysts are forecasting sales of $8.13 and normalized EPS of $4.86, reflecting the exiting of non-core sales, lower profitably from start-up costs and decline sales in its non-auto/industrial related businesses. After the 2023 transition, estimates begin to increase in 2024 (Sales $8.74b and EPS $5.43) through 2025 (Sales $9.65b and EPS $6.39).
The y/y decline in quarterly eps impact from the 2023 transition year appears to be well reflected in analysts’ estimates in the chart below, with an improvement starting in 2024. In Q1, management stated they can meet analysts’ 2023 gross margin expectations.
Revenue estimates paint a similar picture.
What to do from here?
Over the past month, quarterly earnings estimates in 2023 through 2025 are being revised up on the back of its new 2027 financial targets and better than expected Q123 execution, particularly in SiC, which was a catalyst for the recent positive price action.
ON has done a very a good job of guiding the market on a quarterly basis. In the past six quarters, ON has exceeded Wall Street expectations. If this pattern continues in 2023, this could continue to be a positive catalyst for the stock price.
For the remainder of 2023, ON will likely continue to effectively manage Wall Street earnings expectations. While analysts will look for signs that the 2023 transition year is close to reaching the inflection point.
Conclusion
Management deserves credit for implementing smart strategic initiatives and making acquisitions to vertically integrate to capitalize on the secular trends of electrification, automotive ADAS, alternative energy infrastructure and factory automation to position itself as a long term winner.
Although this will impact profitability in 2023, the progression starting in 2024 through 2027 is significant. Importantly, management has done a credible job of explaining the roadmap to analysts and has consistently reached its quarterly targets. Since 2020, the management team has delivered on its strategic and earnings goals.
The valuation does not price in much expectations beyond the 2023 transition year. We expect management to continue its track record of effectively managing wall street quarterly eps expectations for the remainder of the year. Once analysts believe the transition is nearing its completion, likely sometime in 2h23, the focus will shift to 2024 and 2025 earnings potential.
You’ve probably already heard by now that Nvidia’s guidance for fiscal Q2 of $11 billion was 53% higher than analyst expectations of $7.2 billion. The stock is up 25% after hours, adding $200 billion to its market cap.
Originally, the stock was up +15% and then shot up to +25% when the CFO mentioned the beat is coming from the data center, plus when it was confirmed H2 would be also strong with visibility on supply.
We’ve been talking about this rebound for several months. It’s been the subject of quarterly webinars, free editorials and our earnings updates, as well. I was starting to feel like a broken record about the “H2 Semi Rebound” – especially on our last webinar — but I guess it was for a good reason because we got a helluva rebound today.
The rebound now looks like this:
Data center was expected to grow 9% YoY yet beat expectations at 14% YoY growth in the current quarter. This was stated in the opening remarks: “We expect this sequential growth to largely be driven by data center, reflecting a steep increase in demand related to generative AI and large language models. This demand has extended our data center visibility out a few quarters and we have procured substantially higher supply for the second half of the year.”
If Nvidia is adding roughly $4 billion in revenue, primarily driven by the data center, then Q2’s growth will accelerate to an incredible 100% growth rate, up from $3.8 billion in the year ago quarter. It also means the data center will roughly double from the first quarter (sequentially) as the segment was $4.28B in the current quarter.
Put another way, this means Nvidia’s data center segment in the upcoming quarter will be as large as the company’s entire revenue this quarter – if we assume $7.75B in the data center compared to $7.2B total revenue this quarter. See below for analyst Q&A around this.
The other segments are, you guessed it, rebounding too. The sequential numbers show this quite clearly. Here are the official numbers:
Data center revenue of $4.28B up 18% QoQ and 14% YoY
Gaming revenue of $2.2B, up 22% QoQ and down (-38%) YoY
Pro Viz revenue of $295 million, up 31% QoQ and down (-53%) YoY
Automotive revenue of $296 million, up 1% QoQ and up 114% YoY
The gross margin was strong at 64% in the current quarter and the guide is for 68.6% in the next quarter – per the CFO: “Gross margins have now largely recovered to prior peak level, and we have absorbed higher costs, and offset them by innovating and delivering higher valued products as well as products incorporating more and more software.”
Earnings Call:
This question pointed toward the “doubling” of the data center revenue:
C.J. Muse
Yeah. Good afternoon. Thank you for taking the question. I guess with data center, you are essentially doubling quarter-on-quarter, yes, two natural kind of questions that relate to one another come to mind. Number one, where are we in terms of driving acceleration into servers to support AI? And as part of that, as you deal with longer cycle times with TSMC and your other partners, how are you thinking about managing their commitments there with where you want to manage your lead times, in the coming years, to best kind of match — best match that supply and demand? Thanks so much.
Note: the answer was long so this is an excerpt
Jensen Huang:
“And what happened is, when generative AI came along, it triggered a killer app for this computing platform that's been in preparation for some time. And so, now we see ourselves in two simultaneous transitions.The world's $1 trillion data center is nearly populated entirely by CPUs today, and $1 trillion, $250 billion a year, it's growing of course. But over the last four years, call it a $1 trillion worth of infrastructure installed. And it's all completely based on CPUs and dumb NICs (ph). It's basically unaccelerated.
In the future, it's fairly clear now with this — with generative AI becoming the primary workload of most of the world's data centers generating information, it is very clear now that — and the fact that accelerated computing is so energy efficient, that the budget of the data center will shift very dramatically towards accelerated computing and you're seeing that now. We're going through that moment right now as we speak. While the world's data center CapEx budget is limited, but at the same time, we're seeing incredible orders to retool the world's data centers.”
Here was a good question about what Q3 and Q4 will look like if Q2 is this strong, and can supply keep up with the demand:
Vivek Arya:
Thanks for the question. Could I just wanted to clarify does visibility mean data center sales can continue to grow sequentially in Q3 and Q4 or do they sustain at Q2 levels? I just wanted to clarify that. And then Jensen, my question is that, given this very strong demand environment, what does that do to the competitive landscape? Does it invite more competition in terms of custom ASICs? Does it invite more competition in terms of other GPU solutions or other kinds of solutions? How do you see the competitive landscape change over the next two to three years?
Colette Kress:
Yeah, Vivek. Thanks for the question. Let me see if I can add a little bit more color. We believe that the supply that we will have for the second half of the year will be substantially larger than H1. So, we are expecting not only the demand that we just saw in this last quarter, the demand that we have in Q2 for our forecast, but also planning on seeing something in the second half of the year. We just have to be careful here. But we are not here to guide on the second half of that. Yes, we do plan a substantial increase in the second half compared to the first half.
Conclusion:
Nvidia is our largest position and we have taken gains along the way. The plan right now is to wait for a pullback to add more. This earnings report may have changed what the pullback looks like as it’s certainly a historic earnings report in many regards – I do not believe we’ve ever seen $200B market cap added in one day. The reason this much market cap was added AH is that the acceleration of 100% growth in the data center is bonkers considering most tech companies are struggling with a sharp deceleration.
I had put in an editorial that Nvidia investors can take their sweet time adding to this position. This is true, so don’t panic. You haven’t seen anything yet. Once software starts to kick in, it’s going to be an unbelievable ride.
The risk with Nvidia is two things: Broader Macro Events and China. As Jensen Huang recently said: “There is only one China.” If the chip war heats up, China cannot be replaced and the effects would be devastating to the chip industry. Macro continues to be a wild card and tech is especially sensitive to macro. I don’t mean to end on a somber note but I want to make sure we talk about the risks so you’re mentally prepared for it.
You’ve probably already heard by now that Nvidia’s guidance for fiscal Q2 of $11 billion was 53% higher than analyst expectations of $7.2 billion. The stock is up 25% after hours, adding $200 billion to its market cap.
Originally, the stock was up +15% and then shot up to +25% when the CFO mentioned the beat is coming from the data center, plus when it was confirmed H2 would be also strong with visibility on supply.
We’ve been talking about this rebound for several months. It’s been the subject of quarterly webinars, free editorials and our earnings updates, as well. I was starting to feel like a broken record about the “H2 Semi Rebound” – especially on our last webinar — but I guess it was for a good reason because we got a helluva rebound today.
The rebound now looks like this:
Data center was expected to grow 9% YoY yet beat expectations at 14% YoY growth in the current quarter. This was stated in the opening remarks: “We expect this sequential growth to largely be driven by data center, reflecting a steep increase in demand related to generative AI and large language models. This demand has extended our data center visibility out a few quarters and we have procured substantially higher supply for the second half of the year.”
If Nvidia is adding roughly $4 billion in revenue, primarily driven by the data center, then Q2’s growth will accelerate to an incredible 100% growth rate, up from $3.8 billion in the year ago quarter. It also means the data center will roughly double from the first quarter (sequentially) as the segment was $4.28B in the current quarter.
Put another way, this means Nvidia’s data center segment in the upcoming quarter will be as large as the company’s entire revenue this quarter – if we assume $7.75B in the data center compared to $7.2B total revenue this quarter. See below for analyst Q&A around this.
The other segments are, you guessed it, rebounding too. The sequential numbers show this quite clearly. Here are the official numbers:
Data center revenue of $4.28B up 18% QoQ and 14% YoY
Gaming revenue of $2.2B, up 22% QoQ and down (-38%) YoY
Pro Viz revenue of $295 million, up 31% QoQ and down (-53%) YoY
Automotive revenue of $296 million, up 1% QoQ and up 114% YoY
The gross margin was strong at 64% in the current quarter and the guide is for 68.6% in the next quarter – per the CFO: “Gross margins have now largely recovered to prior peak level, and we have absorbed higher costs, and offset them by innovating and delivering higher valued products as well as products incorporating more and more software.”
Earnings Call:
This question pointed toward the “doubling” of the data center revenue:
C.J. Muse
Yeah. Good afternoon. Thank you for taking the question. I guess with data center, you are essentially doubling quarter-on-quarter, yes, two natural kind of questions that relate to one another come to mind. Number one, where are we in terms of driving acceleration into servers to support AI? And as part of that, as you deal with longer cycle times with TSMC and your other partners, how are you thinking about managing their commitments there with where you want to manage your lead times, in the coming years, to best kind of match — best match that supply and demand? Thanks so much.
Note: the answer was long so this is an excerpt
Jensen Huang:
“And what happened is, when generative AI came along, it triggered a killer app for this computing platform that's been in preparation for some time. And so, now we see ourselves in two simultaneous transitions.The world's $1 trillion data center is nearly populated entirely by CPUs today, and $1 trillion, $250 billion a year, it's growing of course. But over the last four years, call it a $1 trillion worth of infrastructure installed. And it's all completely based on CPUs and dumb NICs (ph). It's basically unaccelerated.
In the future, it's fairly clear now with this — with generative AI becoming the primary workload of most of the world's data centers generating information, it is very clear now that — and the fact that accelerated computing is so energy efficient, that the budget of the data center will shift very dramatically towards accelerated computing and you're seeing that now. We're going through that moment right now as we speak. While the world's data center CapEx budget is limited, but at the same time, we're seeing incredible orders to retool the world's data centers.”
Here was a good question about what Q3 and Q4 will look like if Q2 is this strong, and can supply keep up with the demand:
Vivek Arya:
Thanks for the question. Could I just wanted to clarify does visibility mean data center sales can continue to grow sequentially in Q3 and Q4 or do they sustain at Q2 levels? I just wanted to clarify that. And then Jensen, my question is that, given this very strong demand environment, what does that do to the competitive landscape? Does it invite more competition in terms of custom ASICs? Does it invite more competition in terms of other GPU solutions or other kinds of solutions? How do you see the competitive landscape change over the next two to three years?
Colette Kress:
Yeah, Vivek. Thanks for the question. Let me see if I can add a little bit more color. We believe that the supply that we will have for the second half of the year will be substantially larger than H1. So, we are expecting not only the demand that we just saw in this last quarter, the demand that we have in Q2 for our forecast, but also planning on seeing something in the second half of the year. We just have to be careful here. But we are not here to guide on the second half of that. Yes, we do plan a substantial increase in the second half compared to the first half.
Conclusion:
Nvidia is our largest position and we have taken gains along the way. The plan right now is to wait for a pullback to add more. This earnings report may have changed what the pullback looks like as it’s certainly a historic earnings report in many regards – I do not believe we’ve ever seen $200B market cap added in one day. The reason this much market cap was added AH is that the acceleration of 100% growth in the data center is bonkers considering most tech companies are struggling with a sharp deceleration.
I had put in an editorial that Nvidia investors can take their sweet time adding to this position. This is true, so don’t panic. You haven’t seen anything yet. Once software starts to kick in, it’s going to be an unbelievable ride.
The risk with Nvidia is two things: Broader Macro Events and China. As Jensen Huang recently said: “There is only one China.” If the chip war heats up, China cannot be replaced and the effects would be devastating to the chip industry. Macro continues to be a wild card and tech is especially sensitive to macro. I don’t mean to end on a somber note but I want to make sure we talk about the risks so you’re mentally prepared for it.
Feel free to join me on the forum tomorrow for an initial reaction to the earnings report.join me on the forum tomorrow for an initial reaction to the earnings report.
Everyone is expecting Nvidia to selloff soon (including the I/O Fund!). It’ll be interesting to see if popular opinion is proven wrong, only because the market has a way of proving everyone wrong. Call it 2022 leaving a mark to where tech investors are fully prepared to sell now that a stock is up triple digits. But certainly, nothing is wrong with putting money in the bank. We have to perform on an annual basis so clocking gains is something we do frequently and we issue real-time trade alerts to this effect. Knox has included his trading plan below. You may want to handle this position differently due to its strong, long-term potential. In fact, tomorrow I am publishing a free article entitled “Nvidia Will Still Surpass Apple’s Valuation” so that helps frame where I think this company is headed.
If we set aside the AI thesis for a moment, you can see below why Nvidia has rallied as the revenue is expected to rebound from (-21.4%) on for the upcoming quarter to as much as +32% growth by fiscal Q3 ending in September. The popularity of the H100 could lead to a beat somewhere across these next few quarters. In addition, the RTX40 Series lower-end model will be released today for $299 and up, and this may further help the gaming revenue for Q2 and beyond.
Nvidia is unique in that the healthy growth is expected to continue into the foreseeable future — long after the company laps the quarters of the crazy gaming miss. What we don’t want is to invest in companies propped up temporarily by low comps. This is not the case with Nvidia.
Here is Nvidia’s revenue growth over the past five quarters, which shows the effects of the gaming segment:
As stated in our previous earnings coverage on Nvidia, all segments are expected to grow sequentially. I believe this is a major contributor for the rally we are seeing. The market saw what we saw, which was a sharp rebound in the fundamentals and that is critical to understanding why Nvidia is the top stock in the market right now.
Compare that picture to this one:
It’s not only the top line that is rebounding but also the bottom line too (which makes sense but is important to point out):
What’s also important to consider is that Nvidia is expected to grow 20%+ on the top line through mid-2025 and 20%+ on the bottom line through mid-2025.
I will stick my neck out and say that these estimates are too low. Once we start to see AI software contribute to sales and once we add in automotive, Nvidia will be the most hated stock on the market because either people will have tried to short it and gotten burned, or people will have not bought it and will be quite angry with themselves over it.
The H100 is capable of putting up an earnings surprise or two, but most importantly, analysts are not able to truly model AI software yet (at all). Nvidia is going to power every vehicle one day, and drivers will be paying a monthly fee for the software.
That doesn’t mean the company won’t see pullbacks. I’ve included Knox’s most recent TA below.
Moving along, data center revenue was $3.75B in Q1 FY23. The projected mid-point above is $4.075B, representing 8.7% YoY growth and 12.7% growth sequentially. Here is what was said on the call:
“Thanks for the question. First, talking about our data center guidance that we provided for Q1. We do expect a sequential growth in terms of our data center, strong sequential growth. And we are also expecting a growth year-over-year for our data center. We actually expect a great year with our year-over-year growth in data center probably accelerating past Q1.”
So, what we don’t see in the graph above is what the “accelerating past Q1” will be and this is the one data point that can get the stock to move AH.
Gaming will be “meh” this quarter at an expected (-45%) growth – but again, let’s see if we get a decent guide for Q2:
Fiscal Q1 Earnings Consensus:
Here are the nitty gritty numbers for quick reference
Revenue and EPS:
Current quarter revenue of $6.52B for growth of (-21.4%)
Next quarter revenue of $7.08B for growth of +5.7%
Current quarter EPS of $0.92
Next quarter EPS of $1.06
Margins:
Overall, the current quarter margins are higher than FY2023 (ending in Jan) but lower than FY2022. Note, we are currently in FY2024 for Nvidia.
Gross margin guide of 64% is lower than the 57% from FY2023 (ending in Jan)
Adjusted gross margin guide of 66.5% is higher than 59% in FY2023
GAAP Operating margin guide of 25% is higher than 15.7% in FY2023
Adjusted Operating margin guide of 39% is higher than 33.5% in FY2023
Cash Flow:
In the year ago quarter, Nvidia posted $1.73B in op cash flow and $1.35 billion in free cash flow. Margins were 21% and 16.2%, respectively.
Last quarter, Nvidia posted $2.25B in op cash flow and $1.74B in free cash flow. Margins were 37% and 29%, respectively.
The company has $13.3B in cash and $10.95B in debt.
The company returned $1.15B to shareholders in Q4 with $7B remaining for share repurchases
Revenue Segments:
Data center: $3.85B to $4.3B
Gaming: $1.9B to $2.05B
Auto and Pro Viz: $550M to $580M
Additional Notes:
Below are excerpts from an upcoming newsletter that will hit inboxes prior to the earnings report. We have published on the H100 and A100 many times on our premium site but am including this as a quick reference point.
I want to take the opportunity to explain why Nvidia has the ability to arrive at a valuation that is 3X higher than its peers and in some cases 12X higher.
I’m not defending this valuation rather want to explain how it’s possible that smart money continues to buy up here.
The H100 is a Turning Point from Hardware to Software
“The Hopper architecture is ramping and it’s yet again going to disrupt the GPU and AI accelerator market. I’ve written quite a bit about Nvidia […] however, I will keep it simple by saying the A100 GPU is what led the company’s gains since Q2 2020 and the Hopper H100 GPU is what will lead the company’s gains for the next two years.” –Premium Site August 2022, following Nvidia’s $2.5 billion revenue miss, following Nvidia’s $2.5 billion revenue missPremium Site August 2022, following Nvidia’s $2.5 billion revenue miss, following Nvidia’s $2.5 billion revenue miss
Note: the information below is a bit technical, so I’ve bolded the key points for a quick read.
For context, the A100 GPU was a monumental release for Nvidia as the Ampere architecture unified training and inference onto a single chip, whereas in the past Nvidia’s GPUs were mainly used for training.
The result is a 20x performance boost from a multi-instance GPU that allows many GPUs to look like one GPU. The A100 offered the largest leap in performance to date over the past 8 generations. One year later, the Ampere architecture had become the best-selling GPU architecture in the company’s history.
The A100 was special but it’s the H100 that is Nvidia’s iPhone moment. The reason is quite simple – it’s the release that will help Nvidia breakout from hardware and put the company firmly on the map for AI software.
Hardware has allowed Nvidia to become a $700B market cap company, but it is the recurring revenue from AI software that will propel Nvidia into a market cap worth trillions.
You know this story well: the relationship between a hardware company leveraging their position to capture the lion’s share of software — because that’s exactly what Apple did. My contention is that the iPhone was successful because of the moat iOS developers created, and the additional flywheel from the App Store. You can read more here.
The H100 delivers 9X more throughput in AI training, and 16X to 30X more inference performance. The company also states in HPC application-specific workloads, the H100 is 7X faster. The goal of the H100 was not only to add more transistors and make the H100 faster, but to also offer function-specific optimizations. This is achieved through the transformer engine.
The transformer engine is one of the key aspects of the H100. Transformers are becoming one of the most popular neural-network models by applying self-attention to detect how data elements in a series influence and depend on one another.
Prior to transformer models, labeled datasets had to be used to train neural networks. Transformer models eliminate this need by finding patterns between elements mathematically, which substantially opens up what datasets can be used and how quickly. Transformers are partial to the parallel processing that GPUs offer.
The Hopper architecture aims to answer one of the bigger challenges facing superfast compute, which is that moving data into traditional servers overloads the CPU and system memory and becomes bottlenecked by PCI-Express.
By improving the bandwidth issue, Nvidia’s goal is to create more demand for their DGX Pod and SuperPod Systems, which in turn, will create more demand for their software.
The DGX SuperPods scale into a super-GPU capable of 768 terabytes per second. To compare, the entire internet requires 100 terabytes per second. This results in 1 exaflop of FP8 AI performance that runs trillions of parameters. FP8 is most commonly used for inference yet may be used for training in the future due to boosting throughput.
Whereas traditional workloads required many connections exchanging small amounts of data, the workloads of the future will require data to be shared quickly between GPUs and storage. This is accomplished by bypassing the CPU and sending data directly to the GPU while using the network hardware to move the data.
This is ideal for enterprise use cases where people are more likely to use Ethernet while AI and HPC workloads continue to use the Quantum-2 based off Mellanox’s InfiniBand.
Not only will Nvidia begin to monetize through software on the DGX systems but accessibility will improve through CSPs, or cloud service providers. This is an attempt to democratize AI development while driving software sales. On a trailing 4-quarter basis, cloud service providers drove 40% of data center revenue. This is important as cloud service providers will help move DGX Cloud along and AI-as-a-service.
Nvidia’s TAM of $600 Billion is Easily and Quickly Achievable
“The conclusion to my analysis is the same as the introduction, which is that I believe Nvidia is capable of out-performing all five FAAMG stocks and will surpass even Apple’s valuation in the next five years.” – Forbes and Free Newsletter, August 2021Forbes and Free Newsletter, August 2021
Last year, CEO Jensen Huang provided a total addressable market of $300 billion in hardware and $300 billion in software. Meanwhile, Elon Musk is deploying 10,000 GPUs in the cloud and there will likely be tens of thousands more to inference a widely deployed model for a social media generative AI project. Per the analyst on the call: “So it seems like the incremental TAM is easily in the several hundred thousands of GPUs and easily in the tens of billions of dollars. But I'm kind of wondering what this does to the TAM numbers you gave last year. I think you said $300 billion hardware TAM and $300 billion software TAM. So how do you kind of think about what the new TAM would be?”
Huang aptly answered: “I think those numbers are really a good anchor still. The difference is because of the, if you will, incredible capabilities and versatility of generative AI and all of the converging breakthroughs that happened towards the middle and the end of last year, we're probably going to arrive at that TAM sooner than later.”
Today, Nvidia trades at 1.5X this TAM at $773 billion compared to an achievable TAM of $600 billion. This would suggest the stock price does not yet fully reflect the future market opportunity. Also, compare this TAM of $600 billion to Apple’s revenue of $394 billion, which helps illustrate why I said in the past that Nvidia Will Surpass Apple’s Valuation.
What Levels We Plan to Buy Again
Even with logging sizable wins in this position in 2023, it remains our top position while still having enough cash to buy at lower levels.
It is our belief that NVDA is setting up for a sizable pullback, which we believe will open the door for better long-term entries. The reasons for this are below:
The structure/pattern of NVDA’s bounce signals caution. If we look at the pattern off the October low, it may feel like a straight line up; however, you can clearly see 3 swings (marked a, b, c). The first swing up off the low (a), a bearish retrace that makes a higher low (b), and the current swing that we are still in (c).
When we see a 3 wave pattern off of a major low, more times than not, it is a corrective bounce in a larger downtrend. While it may feel impossible at such heights, please keep in mind how sentiment can and does work against us as investors. It felt like tech could never go down in late 2020, and then it felt like it would never go up in Q4 of 2022.
Nvidia is no different, and what we have is a pattern that suggests a larger pullback than most expect is likely, at minimum. So, until this 3 wave pattern can morph into a 5 wave pattern, the odds favor a sizable pullback soon.
Further evidence of this can be seen in how NVDA is now at a significant supply region that marked the top in late 2023. We are now in the region that would constitute a double top playing, and note how price keeps trending higher with less momentum.
We are approaching a double top in conjunction with one of my favorite “sell signals” – when you have price making 3 higher highs, while the momentum indicator being used is making 3 lower highs. This is clearly happening on multiple time frames, which we believe warrants caution.
Similar patterns can be seen on the weekly chart of NVDA below. As price pushes higher, it is doing so on less momentum and less volume. When we see the same pattern on multiple time frames, it further builds the case for caution.
Regarding the targets we are tracking for entries, there are two general paths I see playing out from the price data in the above charts.
The Blue Count suggests that we completed the large degree uptrend that started in 2018. This would put is in a very large corrective rally with the final leg lower coming later this year/early next year. This would have us retest the October lows, and possibly slightly lower. The big tell for this count playing out will be if the coming pullback is a 5 wave pattern pointing down. If we see a large 5 wave drop from the highs, it is signaling that NVDA will likely go lower than most are anticipating.
The Red Count suggests that the October low was THE low. This will still set us up for a sizable pullback into the $220 – $167 region before setting up to make a run to new highs. This count implies that the large uptrend that started in 2018 is not complete and will be targeting fresh highs in the coming year. The tell for this scenario will be if the coming pullback is a 3 wave pattern. If we see a 3 wave pullback, we will look to be heavy buyers in the general target box just outlined.
We issue real-time trade alerts to our Advanced Members so please make sure you are signed up through our Trade Dashboard here to receive our buys, sells, and trims in real-time.Trade Dashboard here to receive our buys, sells, and trims in real-time.
Feel free to join me on the forum tomorrow for an initial reaction to the earnings report.join me on the forum tomorrow for an initial reaction to the earnings report.
Everyone is expecting Nvidia to selloff soon (including the I/O Fund!). It’ll be interesting to see if popular opinion is proven wrong, only because the market has a way of proving everyone wrong. Call it 2022 leaving a mark to where tech investors are fully prepared to sell now that a stock is up triple digits. But certainly, nothing is wrong with putting money in the bank. We have to perform on an annual basis so clocking gains is something we do frequently and we issue real-time trade alerts to this effect. Knox has included his trading plan below. You may want to handle this position differently due to its strong, long-term potential. In fact, tomorrow I am publishing a free article entitled “Nvidia Will Still Surpass Apple’s Valuation” so that helps frame where I think this company is headed.
If we set aside the AI thesis for a moment, you can see below why Nvidia has rallied as the revenue is expected to rebound from (-21.4%) on for the upcoming quarter to as much as +32% growth by fiscal Q3 ending in September. The popularity of the H100 could lead to a beat somewhere across these next few quarters. In addition, the RTX40 Series lower-end model will be released today for $299 and up, and this may further help the gaming revenue for Q2 and beyond.
Nvidia is unique in that the healthy growth is expected to continue into the foreseeable future — long after the company laps the quarters of the crazy gaming miss. What we don’t want is to invest in companies propped up temporarily by low comps. This is not the case with Nvidia.
Here is Nvidia’s revenue growth over the past five quarters, which shows the effects of the gaming segment:
As stated in our previous earnings coverage on Nvidia, all segments are expected to grow sequentially. I believe this is a major contributor for the rally we are seeing. The market saw what we saw, which was a sharp rebound in the fundamentals and that is critical to understanding why Nvidia is the top stock in the market right now.
Compare that picture to this one:
It’s not only the top line that is rebounding but also the bottom line too (which makes sense but is important to point out):
What’s also important to consider is that Nvidia is expected to grow 20%+ on the top line through mid-2025 and 20%+ on the bottom line through mid-2025.
I will stick my neck out and say that these estimates are too low. Once we start to see AI software contribute to sales and once we add in automotive, Nvidia will be the most hated stock on the market because either people will have tried to short it and gotten burned, or people will have not bought it and will be quite angry with themselves over it.
The H100 is capable of putting up an earnings surprise or two, but most importantly, analysts are not able to truly model AI software yet (at all). Nvidia is going to power every vehicle one day, and drivers will be paying a monthly fee for the software.
That doesn’t mean the company won’t see pullbacks. I’ve included Knox’s most recent TA below.
Moving along, data center revenue was $3.75B in Q1 FY23. The projected mid-point above is $4.075B, representing 8.7% YoY growth and 12.7% growth sequentially. Here is what was said on the call:
“Thanks for the question. First, talking about our data center guidance that we provided for Q1. We do expect a sequential growth in terms of our data center, strong sequential growth. And we are also expecting a growth year-over-year for our data center. We actually expect a great year with our year-over-year growth in data center probably accelerating past Q1.”
So, what we don’t see in the graph above is what the “accelerating past Q1” will be and this is the one data point that can get the stock to move AH.
Gaming will be “meh” this quarter at an expected (-45%) growth – but again, let’s see if we get a decent guide for Q2:
Fiscal Q1 Earnings Consensus:
Here are the nitty gritty numbers for quick reference
Revenue and EPS:
Current quarter revenue of $6.52B for growth of (-21.4%)
Next quarter revenue of $7.08B for growth of +5.7%
Current quarter EPS of $0.92
Next quarter EPS of $1.06
Margins:
Overall, the current quarter margins are higher than FY2023 (ending in Jan) but lower than FY2022. Note, we are currently in FY2024 for Nvidia.
Gross margin guide of 64% is lower than the 57% from FY2023 (ending in Jan)
Adjusted gross margin guide of 66.5% is higher than 59% in FY2023
GAAP Operating margin guide of 25% is higher than 15.7% in FY2023
Adjusted Operating margin guide of 39% is higher than 33.5% in FY2023
Cash Flow:
In the year ago quarter, Nvidia posted $1.73B in op cash flow and $1.35 billion in free cash flow. Margins were 21% and 16.2%, respectively.
Last quarter, Nvidia posted $2.25B in op cash flow and $1.74B in free cash flow. Margins were 37% and 29%, respectively.
The company has $13.3B in cash and $10.95B in debt.
The company returned $1.15B to shareholders in Q4 with $7B remaining for share repurchases
Revenue Segments:
Data center: $3.85B to $4.3B
Gaming: $1.9B to $2.05B
Auto and Pro Viz: $550M to $580M
Additional Notes:
Below are excerpts from an upcoming newsletter that will hit inboxes prior to the earnings report. We have published on the H100 and A100 many times on our premium site but am including this as a quick reference point. If you’re looking for more background, I would read this analysis here on AI software which was published for premium members.analysis here on AI software which was published for premium members.
I want to take the opportunity to explain why Nvidia has the ability to arrive at a valuation that is 3X higher than its peers and in some cases 12X higher.
I’m not defending this valuation rather want to explain how it’s possible that smart money continues to buy up here.
The H100 is a Turning Point from Hardware to Software
“The Hopper architecture is ramping and it’s yet again going to disrupt the GPU and AI accelerator market. I’ve written quite a bit about Nvidia […] however, I will keep it simple by saying the A100 GPU is what led the company’s gains since Q2 2020 and the Hopper H100 GPU is what will lead the company’s gains for the next two years.” –Premium Site August 2022, following Nvidia’s $2.5 billion revenue miss, following Nvidia’s $2.5 billion revenue missPremium Site August 2022, following Nvidia’s $2.5 billion revenue miss, following Nvidia’s $2.5 billion revenue miss
Note: the information below is a bit technical, so I’ve bolded the key points for a quick read.
For context, the A100 GPU was a monumental release for Nvidia as the Ampere architecture unified training and inference onto a single chip, whereas in the past Nvidia’s GPUs were mainly used for training.
The result is a 20x performance boost from a multi-instance GPU that allows many GPUs to look like one GPU. The A100 offered the largest leap in performance to date over the past 8 generations. One year later, the Ampere architecture had become the best-selling GPU architecture in the company’s history.
The A100 was special but it’s the H100 that is Nvidia’s iPhone moment. The reason is quite simple – it’s the release that will help Nvidia breakout from hardware and put the company firmly on the map for AI software.
Hardware has allowed Nvidia to become a $700B market cap company, but it is the recurring revenue from AI software that will propel Nvidia into a market cap worth trillions.
You know this story well: the relationship between a hardware company leveraging their position to capture the lion’s share of software — because that’s exactly what Apple did. My contention is that the iPhone was successful because of the moat iOS developers created, and the additional flywheel from the App Store. You can read more here.
The H100 delivers 9X more throughput in AI training, and 16X to 30X more inference performance. The company also states in HPC application-specific workloads, the H100 is 7X faster. The goal of the H100 was not only to add more transistors and make the H100 faster, but to also offer function-specific optimizations. This is achieved through the transformer engine.
The transformer engine is one of the key aspects of the H100. Transformers are becoming one of the most popular neural-network models by applying self-attention to detect how data elements in a series influence and depend on one another.
Prior to transformer models, labeled datasets had to be used to train neural networks. Transformer models eliminate this need by finding patterns between elements mathematically, which substantially opens up what datasets can be used and how quickly. Transformers are partial to the parallel processing that GPUs offer.
The Hopper architecture aims to answer one of the bigger challenges facing superfast compute, which is that moving data into traditional servers overloads the CPU and system memory and becomes bottlenecked by PCI-Express.
By improving the bandwidth issue, Nvidia’s goal is to create more demand for their DGX Pod and SuperPod Systems, which in turn, will create more demand for their software.
The DGX SuperPods scale into a super-GPU capable of 768 terabytes per second. To compare, the entire internet requires 100 terabytes per second. This results in 1 exaflop of FP8 AI performance that runs trillions of parameters. FP8 is most commonly used for inference yet may be used for training in the future due to boosting throughput.
Whereas traditional workloads required many connections exchanging small amounts of data, the workloads of the future will require data to be shared quickly between GPUs and storage. This is accomplished by bypassing the CPU and sending data directly to the GPU while using the network hardware to move the data.
This is ideal for enterprise use cases where people are more likely to use Ethernet while AI and HPC workloads continue to use the Quantum-2 based off Mellanox’s InfiniBand.
Not only will Nvidia begin to monetize through software on the DGX systems but accessibility will improve through CSPs, or cloud service providers. This is an attempt to democratize AI development while driving software sales. On a trailing 4-quarter basis, cloud service providers drove 40% of data center revenue. This is important as cloud service providers will help move DGX Cloud along and AI-as-a-service.
Nvidia’s TAM of $600 Billion is Easily and Quickly Achievable
“The conclusion to my analysis is the same as the introduction, which is that I believe Nvidia is capable of out-performing all five FAAMG stocks and will surpass even Apple’s valuation in the next five years.” – Forbes and Free Newsletter, August 2021Forbes and Free Newsletter, August 2021
Last year, CEO Jensen Huang provided a total addressable market of $300 billion in hardware and $300 billion in software. Meanwhile, Elon Musk is deploying 10,000 GPUs in the cloud and there will likely be tens of thousands more to inference a widely deployed model for a social media generative AI project. Per the analyst on the call: “So it seems like the incremental TAM is easily in the several hundred thousands of GPUs and easily in the tens of billions of dollars. But I'm kind of wondering what this does to the TAM numbers you gave last year. I think you said $300 billion hardware TAM and $300 billion software TAM. So how do you kind of think about what the new TAM would be?”
Huang aptly answered: “I think those numbers are really a good anchor still. The difference is because of the, if you will, incredible capabilities and versatility of generative AI and all of the converging breakthroughs that happened towards the middle and the end of last year, we're probably going to arrive at that TAM sooner than later.”
Today, Nvidia trades at 1.5X this TAM at $773 billion compared to an achievable TAM of $600 billion. This would suggest the stock price does not yet fully reflect the future market opportunity. Also, compare this TAM of $600 billion to Apple’s revenue of $394 billion, which helps illustrate why I said in the past that Nvidia Will Surpass Apple’s Valuation.
What Levels We Plan to Buy Again
Even with logging sizable wins in this position in 2023, it remains our top position while still having enough cash to buy at lower levels.
It is our belief that NVDA is setting up for a sizable pullback, which we believe will open the door for better long-term entries. The reasons for this are below:
The structure/pattern of NVDA’s bounce signals caution. If we look at the pattern off the October low, it may feel like a straight line up; however, you can clearly see 3 swings (marked a, b, c). The first swing up off the low (a), a bearish retrace that makes a higher low (b), and the current swing that we are still in (c).
When we see a 3 wave pattern off of a major low, more times than not, it is a corrective bounce in a larger downtrend. While it may feel impossible at such heights, please keep in mind how sentiment can and does work against us as investors. It felt like tech could never go down in late 2020, and then it felt like it would never go up in Q4 of 2022.
Nvidia is no different, and what we have is a pattern that suggests a larger pullback than most expect is likely, at minimum. So, until this 3 wave pattern can morph into a 5 wave pattern, the odds favor a sizable pullback soon.
Further evidence of this can be seen in how NVDA is now at a significant supply region that marked the top in late 2023. We are now in the region that would constitute a double top playing, and note how price keeps trending higher with less momentum.
We are approaching a double top in conjunction with one of my favorite “sell signals” – when you have price making 3 higher highs, while the momentum indicator being used is making 3 lower highs. This is clearly happening on multiple time frames, which we believe warrants caution.
Similar patterns can be seen on the weekly chart of NVDA below. As price pushes higher, it is doing so on less momentum and less volume. When we see the same pattern on multiple time frames, it further builds the case for caution.
Regarding the targets we are tracking for entries, there are two general paths I see playing out from the price data in the above charts.
The Blue Count suggests that we completed the large degree uptrend that started in 2018. This would put is in a very large corrective rally with the final leg lower coming later this year/early next year. This would have us retest the October lows, and possibly slightly lower. The big tell for this count playing out will be if the coming pullback is a 5 wave pattern pointing down. If we see a large 5 wave drop from the highs, it is signaling that NVDA will likely go lower than most are anticipating.
The Red Count suggests that the October low was THE low. This will still set us up for a sizable pullback into the $220 – $167 region before setting up to make a run to new highs. This count implies that the large uptrend that started in 2018 is not complete and will be targeting fresh highs in the coming year. The tell for this scenario will be if the coming pullback is a 3 wave pattern. If we see a 3 wave pullback, we will look to be heavy buyers in the general target box just outlined.
We issue real-time trade alerts to our Advanced Members so please make sure you are signed up through our Trade Dashboard here to receive our buys, sells, and trims in real-time.Trade Dashboard here to receive our buys, sells, and trims in real-time.
Super Micro, also known as Supermicro, is sandwiched in the AI trend between hyperscalers and major chip design companies. The company is a server maker that started off by making motherboards and other components before it began making complete systems. The company is unique in that it sits between being an equipment manufacturer (Dell, HP) and being a design manufacturer (Foxconn).
The company competes with Dell, IBM, Hewlett Packard, and China’s Inspur. The chart below gives you a general idea of the landscape although Supermicro has doubled its revenue since 2021. The server maker is now a $6.8 billion company and ended 2021 as a $3.5B company (Counterpoint estimates were about $1B off) but the chart below is useful in visualizing the competitors.
The word “competitor” is being used loosely here as many of these companies will not necessarily be able to compete on liquid cooling for AI development platforms, or with Supermicro’s Building Block design. The companies pictured above have stagnated and this has worked out for Supermicro, a company that could have stagnated but continued to innovate instead.
According to IDC, the worldwide server market forecast is expected to deceleratedecelerate from 20% to 0%. Supermicro declined in revenue going into 2023, however, according to management, this is due to supply chain issues. This is important to distinguish as the 2023 bull case for Supermicro rests on the high demand the company is seeing, that due to supply chain issues, the company is unable to fill. The prevailing bull thesis is that Supermicro’s supply chain issues will ease in the near term.
According to the most recent Investor’s Presentation, Supermicro grew 5X faster compared to the industry average for subsystems and server systems. While there is pressure for tech management teams to cut costs, Supermicro may be more insulated by working closely on making AI systems with the most cutting-edge chips. This includes AMD 4th Generation Zen Epyc processors, Intel Xeon and soon Sapphire Rapids, Nvidia Grace CPUs and Ampere Arm-based third-gen CPUs, Nvidia’s A100 and H100s GPUs, AMD’s MI250 and MI300 GPUs and Intel’s Ponte Vecchio GPUs.
Another piece to the bull case for Supermicro is the near-term goal to reach $10 billion, which will put the company behind Dell and Hewlett Packard. Should the company reach its $20 billion long-term goal, then it could very well be the leader or at least a strong rival to Dell and HP. If/When this happens, it’ll be due to AI systems. It was stated on the call that AI/GPU and rack-scale solutions represented 29% of our total revenue and the company expects “significant future growth.”
Supermicro’s revenue quickly accelerated last year due to one large customer in Q3 2022 to Q1 2023. This one large customer, which was later identified as Meta, accounted for upward of 20% of revenue in June and September of 2022. By December, Meta had accounted for 10% of revenue. This was the subject of a short report. However, if you invest in a small cap or low mid-cap semiconductor company, there is going to be high customer concentration.
There was also a hint on the call that another customer may be ramping: “an existing Cloud Service Provider customer represented more than 10% of revenues for the first time.”
Liquid Cooling
As the performance of CPUs and GPUs increase, the heat these systems generate increases. Liquid cooling is becoming a popular alternative to air cooling to sustain maximum performance with the added benefit of driving down costs for supercomputers. According to a press release in 2021, liquid cooling can improve data center power usage effectiveness (PUE) and total cost of ownership (TCO) “by over 40% on power costs.”
Here's a quote from the CEO on the importance of this competitive advantage:
“The power consumption and thermal challenges of these new technologies have risen dramatically and 40KW or even 80KW rack solution demands are getting stronger and popular for computing hungry DC and industries. Having high power efficiency and air/liquid thermal expertise has become one of our key differentiators of success.”Having high power efficiency and air/liquid thermal expertise has become one of our key differentiators of success.”
In 2022, Supermicro stated that liquid cooling is being used in 10% of supercomputers but will grow to be used in the “vast majority” in order to offset the heat generated by power-consuming components. The company offers Direct to Chip cooling, Immersion cooling and Rear-door Heat Exchanger cooling. This design works better than air cooling, which needs air conditioning and server fans to run constantly.
Direct to Chip Cooling: Running a cold liquid over the top of a running chip by using a pump to circulate liquid. This is a closed loop system, or also known as a self-contained cooling system.
Immersion Cooling: The system is immersed in liquid for cooling.
Rear Door Heat Exchanger: Uses a specialized rear door to the rack where coolant absorbs the heat.
Water removes heat better than air. Liquid molecules are closer together than air molecules, which results in higher heat transfer. Artificial intelligence/Machine Learning and Big Data require massive amounts of data processing, and as future generations of CPUs and GPUs are released, these systems will exceed air cooling capacity.
Liquid cooling also solves CPU throttling, which occurs when CPUs and GPUs overheat and are throttled back to avoid damage to the chip.
AI Development Platforms
AI development platforms remove the need for disparate hardware and software by offering an end-to-end platform. Supermicro has partnered with Nvidia to offer Certified systems with the new H100 GPUs for the Nvidia AI Enterprise Platform.
These systems come with 3-year AI enterprise software subscriptions, and include workflows, frameworks, pretrained models and infrastructure optimization, in the cloud, in the data center and at the edge.
Supermicro is closely partnered with Nvidia on the H100 GPU rollout with air flow designs that reduce fan speeds, lower power consumption, lower noise levels and lower the total cost of ownership.
The most recent system announced in March enables AI workloads to be run in offices and the system can be rack-mounted, as well, for data center environments. The self-contained cooling system reduces operating expenses and helps the machine to run quietly for AI, deep learning, machine learning and high-performance computing (HPC) applications.
Supermicro is able to deliver workload optimized products quickly because of its building block design. The AI market is moving quickly, and SMCI can build and validate systems partly due to a modular design that allows systems to be updated from new products, such as when Nvidia, AMD or Intel have new design releases.
Financials:
Supermicro saw strong price action due to strong guidance for next quarter and due to strong guidance of 20% revenue growth for fiscal year 2024, which begins in July.
The current quarter revenue and EPS missed management guidance and analyst estimates. SMCI reported $1.28 billion for fiscal Q3 growth of (5%), which missed guidance of $1.48 billion, at the midpoint. GAAP EPS was $1.53 and Adjusted EPS was $1.71. This compares to management guidance of: “GAAP diluted net income per share of $1.75 to $2.02 and non-GAAP diluted net income per share of $1.88 to $2.14.”
This was Supermicro’s first miss dating back to 2019. Management said the following about the miss: “The shortfall was primarily due to key new component shortages for Supermicro’s new generation server platforms which have been mostly resolved to-date.”
The fiscal Q4 revenue guide was for $1.7B to $1.9B, which is above the $1.64B that analysts were expecting. The EPS forecast form management is for $2.21 to $2.71, compared to analyst expectations of $1.76.
Per the earnings call: “If supply conditions improve sooner, we expect to be above that range, despite some economic headwinds ahead. In other words, I continue to expect our fiscal year 2024 revenue to be at least 20% year-over-year growth and we are accelerating to reach our mid- to-long-term growth objectives of $20 billion per year.”
According to a previous earnings call, the CFO stated: “GPU prices and CPU prices are going up, especially with the new refreshes that are coming out. So we anticipate that [average selling prices] will continue to go up.”
At 20% growth, the company will surpass $8 billion in revenue next fiscal year, ending in June. The FY2025 analyst estimates are for growth of 11%.
Margins:
Server solutions and systems come with thin gross margins. Despite thin margins, Supermicro is a company with strong earnings with EPS of $10.73 for FY2023. Please see below for questions from analysts on the call regarding gross margin.
Gross Margin of 17.6% compares to 15.5% in the year ago quarter. The company stated GM was lower due to product mix and new platform ramps.
GAAP operating margin of 7.7% compares to 6.6% in the year ago quarter. This is lower compared to previous quarters this year in the 9% to 10% range. The company stated it was lower due to lower revenue.
GAAP net margin of 6.7% is up from 5.7% a year ago. This is lower compared to previous quarters this year in the 9% to 10% range.
Cash Flow:
The company reported strong cash flow in the current quarter with a margin of 15.5% for operating cash flow and 14.8% margin on free cash flow. This equals $198 million and $190 million, respectively. There is $362 million on the balance sheet and $187 million in debt.
Notably, the company has lumpy free cash flow with FY2022 and FY2019 ending negative. Here’s a snapshot of the most recent quarters:
According to management, Supermicro’s decline in revenue growth is due to supply chain issues. Per the CFO’s opening remarks:
“Fiscal Q3, 2023 revenues were $1.28 billion, down 5% year-over-year and down 29% quarter-over-quarter, which was below our initial guidance range of $1.42 billion to $1.52 billion. The shortfall was primarily due to key new component shortages for Supermicro’s new generation server platforms which have been mostly resolved to-date.
We note that our shipments against a record backlog may be constrained by supply chain bottlenecks due to high demand for our advanced AI server platforms.”
Supermicro builds complete systems, and the supply chain issues can extend beyond CPUs, GPUs and memory to also include difficulty obtaining metal-oxide-semiconductor field-effect transistors (MOSFETs), diodes and capacitors. If there is low availability with these components, the systems won’t be complete in order to ship. According to The Register, lead times were at 26 weeks back in October compared to the 10-14 weeks that is the target for a healthy supply chain. This is an improvement from 40 weeks a year ago.
Here was another comment on the earnings call:
“These new AI product demands from top-tier companies have led us to challenges in terms of new key components availability.
Compounded with the economic headwind, our Q3 results were reflective of these difficult yet opportune conditions. The good news is that we have already started to address these component shortage pressures over the past few months and we are in a much-improved situation going forward. We have started to produce and ship some back orders since April.”
Risks:
Investors risk entering a frothy AI market. Most tech investors have mastered the extreme exuberance followed by extreme fear that tech seems to oscillate between. We are nearing extreme exuberance on AI with social media exploding over Chat-GPT and Bard. It’s rarely a good sign, and I’m saying that as someone who is exposed to AI-related stocks and benefits from this exuberance. Entering AI stocks right now should be done with a stop in mind.
This company had a short report out earlier this year that caused the stock to selloff. You can read the concerns here.
Regarding fundamentals, the gross margin is the primary concern raised in the earnings calls.
Yeah. So, Nehal, we — yeah. Back two years ago, we gave a 17% to 21% — 23% topline growth. Obviously, we’re in there at a minimum of 20%. And for the gross margins, we continue to, like I said, to wrestle with taking market share and also balancing that against gross margins.
But we’re confident with our new manufacturing facilities coming online that we will be able to improve our gross margins. And we also, as we come out of this quarter and we begin to ramp our new product offerings that we will be able to improve margins as well.
Here was another question about the gross margin:
Ananda Baruah:
But I would love to get a better understanding of how you guys are thinking about sort of the gross margin manifestation if we think about the continued layering in of larger footprint, which may come at a slightly lower margin. Is it really that over time, we just expect a greater presence of that lower margin business with some efficiency gains or is it just in the beginning here, the margin will be lower for the new business, but then collectively, the P&L gross margin expands over time?
David Weigand:
Yeah. So we’re looking at it and on — in the — your latter alternative, Ananda, and here’s why. So right now, there’s three things that we’ve been facing. We’re having to face more air transportation costs in order to make our deliveries. So that impacts our margin. And also, we’re having to pay other expedite fees. That impacts our margin.
Number two, we ran a lot less through our factories than in Q3 than we did in Q2. So your margin efficiency, your ability to spread your fixed costs, it’s tremendously impacted on a smaller scale. So as we scale up, we improve our margins.
Thirdly, the — as we ramped our new product offerings, there is an efficiency on these new — on the production of these new products. So we are going to improve the efficiency of these products, which will improve the margin. And so those three things alone speak to margin improvements.
Charles Liang
I can add some color. I mean, as I shared, I mean, we are building a $20 billion of revenue, hopefully in midterm and that’s why a grow our capacity and support a large customer is very important to us. Once our volume becomes higher, our costs will be improved and then business operation efficiency will be higher.
So we are doing better great way to grow our revenue. And so, I mean, once we start to reach that number under $10 billion to $20 billion, I guess, our gross margin will start to grow, because we won’t always invest for big growth after that.
Valuation:
Supermicro has an old school semiconductor top line valuation that reflects its roots as a server maker. Interesting enough, it’s trading at its previous 2015 high. The 5-year median is 0.46 but there’s been too big of a product pivot to rely on this for the future valuation.
On the bottom line, SMCI trades in line with Intel. The bottom line is probably a better gauge for this stock than top line. I took a screenshot with July 2022 marked so you could see where AMD typically trades when it’s not going through a major cyclical event with PCs. It also shows where Intel and SMCI were back in July versus now – about 50% higher valuations.
The 5-year median for SMCI is 15 and has been up to a 20 5-year median in the past.
Conclusion:
The AI market is frothy but we may take a shot at entering. If we stop out, then no big deal. We’d like exposure to Supermicro for its ambitions to overtake the incumbents, plus the clear path the company will take to do so.
Please note, this was originally published for Premium Members on May 2nd when the stock was down (9%) and prior to the Microsoft announcement. This helps to illustrate our conviction in the face of a fickle market reaction, as two days later, AMD’s AI strategy became clearer with the Microsoft announcement and the stock was up quite a bit. Also, due to a tight earnings schedule for our portfolio and pipeline, you can expect to receive the May stock tip before the end of this week or Monday at the latest.Microsoft announcement. This helps to illustrate our conviction in the face of a fickle market reaction, as two days later, AMD’s AI strategy became clearer with the Microsoft announcement and the stock was up quite a bit. Also, due to a tight earnings schedule for our portfolio and pipeline, you can expect to receive the May stock tip before the end of this week or Monday at the latest.
Back in the Fall, when Nvidia was badly beaten up, I wrote two editorials about Nvidia’s gaming bottom and how the company was “ready to rumble” with H100 GPUs. I could repeat these editorials and simply replace the headlines by saying AMD is bottoming on PCs and that AMD is “ready to rumble” with Genoa, Bergamo — and most especially, the highly anticipated MI300 GPUs. Considering Nvidia was down roughly 60% when I wrote those articles, AMD being up roughly 50% YTD doesn’t feel so contrarian – but it is a bit contrarian because the company is on the precipice of proving it’s up for the task of AI acceleration.
The product road map for AMD is particularly exciting right now and also a bit complex because AMD’s AI ambitions are also found outside the data center. We’ve focused on AMD taking market share from Intel for three years with EPYC processors, but now we need to switch our focus and prepare for the next three years. You can expect a fresh deep dive on AMD come this June that drills into this particular company’s AI opportunity. It’ll be the end of a chapter on our site to place our focus beyond Intel for AMD’s growth, but now is the right time so we can prepare for AMD’s moment in AI.
Q1 Highlights/Lowlights:
Data center growth declined 23% sequentially and was flat year-over-year. Per the conversations on the call, data center growth is expected to be up year-over-year and this implies a 50% growth rate in H2 (see below). The sequential decline this quarter is a reflection of enterprise sales and not from EPYC CPUs, which remained strong.
The Client segment was down (65%) due to PCs. The strategy has been to undership for a quicker rebound, which was Nvidia’s strategy for gaming. This makes it more painful in the short term but sets up a better recovery in the long term. According to management, this is the bottom for PCs with a meaningful rebound in H2.
For comparison, we covered this strategy to undership gaming units for Nvidia in the Q2 August earnings report when we wrote:
The CFO Collette Kress stated: “Across those two quarters, the Q2 of ‘23, the Q3 of ‘23, we have likely undershipped gaming to our end demand significantly. We expect that sell-through or essentially our end demand for those combined two quarters of Q2 and Q3 to be approximately $5 billion […].”We expect that sell-through or essentially our end demand for those combined two quarters of Q2 and Q3 to be approximately $5 billion […].”
She is referring to about $1 billion being under shipped (or reduced sell-in) if we assume flat growth for gaming next quarter as the company attempts to rebalance inventory. It would be even more of an under shipment if gaming does decline sequentially.
Gaming was down (6%) but is expected to benefit from the Radeon 7000 Series release. However, gaming is expected to modestly decline again next quarter.
Embedded remains strong with 60% growth and this is part of what our deep dive will discuss as AMD’s product road map on AI is broader than the data center. Notably, coming off strong growth for 4-5 quarters, next quarter Embedded is expected to report a modest decline.
Margins are weaker than usual and this is due to the Client segment/PCs. When product mix returns to normal, margins will stabilize again. Notably, the data center segment had a lower margin of 11% compared to 33% last year. This was due to the Pensando acquisition and AI investments in GPUs.
AMD beat on revenue and EPS in the current March quarter yet revenue missed consensus for Q2 ending in June.
Revenue for Q1 was $5.4 billion compared to $5.3 billion expected. Guidance of $5.3B missed consensus of $5.52B.
EPS beat at $0.60 versus $0.56 consensus.
Margins are lower than usual but came in as expected. Per the note above and the earnings call discussion below, margins will stabilize when PCs return to growth. This is expected to be the bottom for the PC slump.
GAAP Gross Margin is soft at 44% compared to 48% in the year ago quarter. Adjusted gross margin has been steady at 50% and is expected to be 50% again next quarter.
It’s better to focus on adjusted operating margin for now as GAAP operating margin includes expenses from Xilinx and Pensando acquisitions. Adjusted operating margin of 21% was weaker than 31% in the year ago quarter.
Cash flow margins are down with FCF margin of 6% compared to 16% a year ago. Cash and debt has not changed at $5.9 billion and $2.5 billion, respectively.
The company returned $241 million to shareholders through share repurchases. There is $6.3 billion in remaining authorization for share repurchases.
Earnings Call:
Data Center Growth in H2:
There were two important discussions around what data center growth will be in the second half of the year. The first analyst believed the growth would be in the 30% range but it was later confirmed it’ll be in the 50% year-over-year range. AMD investors need the data center to participate in a big way in the second half, therefore both discussions are important, and are highlighted below.
Vivek Arya:
For my first one, Lisa, when I look at your full year Data Center outlook for some growth, that implicitly suggest Data Center, right, could be up 30% in the second half versus the first half, right? And I'm curious, what is your confidence and visibility and some of the assumptions that go into that view?
Is it — do you think there is a much bigger ramp in the new products? Is it enterprise recovery? Is it pricing? So just give us a sense for how we should think about the confidence and visibility of the strong ramp that is implied in your second half Data Center outlook?
Lisa Su
Right. So Vivek, thanks for the question. Maybe let me give you some context on what's going on in the Data Center right now. First of all, we have said that it's a mixed environment in the Data Center. So the first half of the year, there are some of the larger cloud customers that are working through some inventory and optimization as well as a weaker enterprise.
As we go into the second half of the year, we see a couple of things. First, our road map is very strong. So the feedback that we're getting working with our customers on, it's ramping well. It is very differentiated in terms of TCO and overall performance. So we think it's very well positioned. Much of the work that we've done in the first half of the year — in the first quarter and here in the second quarter is to ensure that we complete all of that work such that we can ramp across a broader set of workloads as we go into the second half of the year.
And then I would say, from an overall market standpoint, I think enterprise will still be mixed with the notion that we expect some improvement. It depends a little bit on the macro situation. And then as we go into the second half of the year, in addition to Genoa, we're also ramping Bergamo. So that's on track to launch here in the second quarter and will ramp in the second half of the year. And then as we get towards the end of the year, we also have our GPU ramp of MI300. So with that, we start the ramp in the fourth quarter of our supercomputing wins as well as our early cloud AI wins. So those are all the factors. Of course, we'll have to see how the the year play about how we're positioned from an overall product and road map standpoint for Data Center.”
Below, is where the year-over-year growth rate was confirmed:
“Stacy Rasgon
For my first one, Lisa, can you just like clarify this explicitly for me. So you said double-digit Data Center. Was that a full year statement? Or was that a second half year-over-year statement? Or was that a half-over-half statement for Data Center?
Lisa Su
Yes. Let me be clear. That was a year-over-year statement. So double-digit Data Center growth for the full year of 2023 versus 2022.
Stacy Rasgon
Got it. Which just given what you did in Q1 and sort of are implying for Q2 needs something like 50% year-over-year growth in the second half to get there.So you're endorsing those — you're endorsing that now?
Lisa Su
I am…
Jean Hu
Yes, your math is right.”
Comments on the Client Segment Stabilizing in H2:
Arguably, how the Client segment performs is equally important to how data center performs simply because this segment has been a bit awful for a wide range of companies.
Per the opening remarks:
“Now turning to our client segment. Revenue declined 65% year-over-year to $739 million as we shipped significantly below consumption to reduce downstream inventory. As we stated on our last earnings call, we believe the first quarter was the bottom for our client processor business. We expanded our leadership desktop and notebook processor portfolio significantly in the quarter. In desktops, we launched the industry's fastest gaming processors with our Ryzen 7000 X3D series CPUs that combine our Zen 4 core with industry-leading 3D chiplet packaging technology.”
There were questions in the Q&A with the CEO, Lisa Su, saying the following:
“Lisa Su
Yes. So we've been undershipping sort of consumption in the Client business for about 3 quarters now. And certainly, our goal has been to normalize the inventory in the supply chain so that shipments would be closer to consumption. We expect that, that will happen in the second half of the year, and that's what the comment meant that we believe that there will be improvements in the overall inventory positioning.
And then we also believe that the Client market is stabilizing. So Q1 was the bottom for our business as well as for the overall market. From what we see although it will be a gradual set of improvements, we do see that the overall market should be better in the second half of the year. We like our product portfolio a lot. I'm excited about having AI-enabled on our Ryzen 7000 series. And we have leadership notebook platforms with Dragon Range. Our desktop road map is also quite strong with our new launch of the Verizon 7000 X3D products.
And so I think here in the second quarter, we'll still undership consumption a bit. And by the second half of the year, we should be more normalized between shipments and consumption, and we expect some seasonal improvement into the second half.”
Comments on the MI300 Release:
The MI300 will be released in Q4 and is expected to contribute to revenue by early 2024. I will expand on this in the upcoming deep dive. Here is what was said in the opening remarks:
“Customer interest has increased significantly for our next-generation instinct MI300 GPUs for both AI training and inference of large language models. We made excellent progress achieving key MI300 silicon and software readiness milestones in the quarter, and we're on track to launch MI300 later this year to support the El Capitan exascale supercomputer win at Lawrence Livermore National Laboratory and large cloud AI customers.”
There were many questions about the MI300, one of the more important one being from Ross Seymore who hints toward Nvidia being hard to compete against. Lisa Su’s answer points toward potentially undercutting Nvidia on price and also having strong relationships from EPYC CPUs:
Ross Seymore
And pivoting for my follow-up on the AI side and MI300. I just wanted to know what you would describe as your competitive advantages? Everybody knows that's a market that's exploding right now. There's tons of demand. You guys have all the IP to be able to attack it. But there's a very large incumbent in that space as well.
So when you think about what AMD can bring to the market, whether it's hardware, software, heterogeneity of the products you can bring, et cetera, what do you think is the core competitive advantage that can allow you to penetrate that market successfully?
Lisa Su
Yes. There's a couple of aspects, Ross. And — yes, since we haven't yet announced MI300, all of the specifications will — some of those will come over the coming quarters. MI300 is the first solution that has both the CPU and GPU together, and that has been very positive for the supercomputing market.
I think as it relates to generative AI, and we think we have a very strong value proposition from both a hardware and again, it's a performance per dollar conversation, I think there's a lot of demand in the market. And there's also — I think given our deep customer relationships on the EPYC side, there's actually a lot of synergy between the customer set between the EPYC CPUs and the sort of 300 GPU customers.
Conclusion:
This is the kind of earnings report where you’d have to listen to the call to piece it together. Journalists and others who cover dozens of stocks will rush the conclusion (beat this quarter, lower guidance quarter) but there a lot of depth to AMD right now – perhaps more depth than any other stock we own right now since the product road map is loaded for the second half.
I’m still feeling Zen after this earnings report and also feeling like it’s time to dedicate a new deep dive to the company. It’ll be a bit nostalgic to come back to Xilinx again as I called the company “a heavy hitter AI stock” four years ago. The MI300 GPUs are exciting but Xilinx is also a large piece to AMD’s AI opportunity. Keep an eye out for this deep dive in the next couple weeks as we carefully, patiently, and with an exceptional level of due diligence, build out the world’s very best AI portfolio.
Deep dives, trade alerts, a forum and weekly webinars on the I/O Fund portfolio are offered on our premium service, you can find out more information here.here.
Back in the Fall, when Nvidia was badly beaten up, I wrote two editorials about Nvidia’s gaming bottom and how the company was “ready to rumble” with H100 GPUs. I could repeat these editorials and simply replace the headlines by saying AMD is bottoming on PCs and that AMD is “ready to rumble” with Genoa, Bergamo — and most especially, the highly anticipated MI300 GPUs. Considering Nvidia was down roughly 60% when I wrote those articles, AMD being up roughly 50% YTD doesn’t feel so contrarian – but it is a bit contrarian because the company is on the precipice of proving it’s up for the task of AI acceleration.
The product road map for AMD is particularly exciting right now and also a bit complex because AMD’s AI ambitions are also found outside the data center. We’ve focused on AMD taking market share from Intel for three years with EPYC processors, but now we need to switch our focus and prepare for the next three years. You can expect a fresh deep dive on AMD come this June that drills into this particular company’s AI opportunity. It’ll be the end of a chapter on our site to place our focus beyond Intel for AMD’s growth, but now is the right time so we can prepare for AMD’s moment in AI.
Q1 Highlights/Lowlights:
Data center growth declined 23% sequentially and was flat year-over-year. Per the conversations on the call, data center growth is expected to be up year-over-year and this implies a 50% growth rate in H2 (see below). The sequential decline this quarter is a reflection of enterprise sales and not from EPYC CPUs, which remained strong.
The Client segment was down (65%) due to PCs. The strategy has been to undership for a quicker rebound, which was Nvidia’s strategy for gaming. This makes it more painful in the short term but sets up a better recovery in the long term. According to management, this is the bottom for PCs with a meaningful rebound in H2. For comparison, we covered this strategy to undership gaming units for Nvidia here.
Gaming was down (6%) but is expected to benefit from the Radeon 7000 Series release. However, gaming is expected to modestly decline again next quarter.
Embedded remains strong with 60% growth and this is part of what our deep dive will discuss as AMD’s product road map on AI is broader than the data center. Notably, coming off strong growth for 4-5 quarters, next quarter Embedded is expected to report a modest decline.
Margins are weaker than usual and this is due to the Client segment/PCs. When product mix returns to normal, margins will stabilize again. Notably, the data center segment had a lower margin of 11% compared to 33% last year. This was due to the Pensando acquisition and AI investments in GPUs.
AMD beat on revenue and EPS in the current March quarter yet revenue missed consensus for Q2 ending in June.
Revenue for Q1 was $5.4 billion compared to $5.3 billion expected. Guidance of $5.3B missed consensus of $5.52B.
EPS beat at $0.60 versus $0.56 consensus.
Margins are lower than usual but came in as expected. Per the note above and the earnings call discussion below, margins will stabilize when PCs return to growth. This is expected to be the bottom for the PC slump.
GAAP Gross Margin is soft at 44% compared to 48% in the year ago quarter. Adjusted gross margin has been steady at 50% and is expected to be 50% again next quarter.
It’s better to focus on adjusted operating margin for now as GAAP operating margin includes expenses from Xilinx and Pensando acquisitions. Adjusted operating margin of 21% was weaker than 31% in the year ago quarter.
Cash flow margins are down with FCF margin of 6% compared to 16% a year ago. Cash and debt has not changed at $5.9 billion and $2.5 billion, respectively.
The company returned $241 million to shareholders through share repurchases. There is $6.3 billion in remaining authorization for share repurchases.
Earnings Call:
Data Center Growth in H2:
There were two important discussions around what data center growth will be in the second half of the year. The first analyst believed the growth would be in the 30% range but it was later confirmed it’ll be in the 50% year-over-year range. AMD investors need the data center to participate in a big way in the second half, therefore both discussions are important, and are highlighted below.
Vivek Arya:
For my first one, Lisa, when I look at your full year Data Center outlook for some growth, that implicitly suggest Data Center, right, could be up 30% in the second half versus the first half, right? And I'm curious, what is your confidence and visibility and some of the assumptions that go into that view?
Is it — do you think there is a much bigger ramp in the new products? Is it enterprise recovery? Is it pricing? So just give us a sense for how we should think about the confidence and visibility of the strong ramp that is implied in your second half Data Center outlook?
Lisa Su
Right. So Vivek, thanks for the question. Maybe let me give you some context on what's going on in the Data Center right now. First of all, we have said that it's a mixed environment in the Data Center. So the first half of the year, there are some of the larger cloud customers that are working through some inventory and optimization as well as a weaker enterprise.
As we go into the second half of the year, we see a couple of things. First, our road map is very strong. So the feedback that we're getting working with our customers on, it's ramping well. It is very differentiated in terms of TCO and overall performance. So we think it's very well positioned. Much of the work that we've done in the first half of the year — in the first quarter and here in the second quarter is to ensure that we complete all of that work such that we can ramp across a broader set of workloads as we go into the second half of the year.
And then I would say, from an overall market standpoint, I think enterprise will still be mixed with the notion that we expect some improvement. It depends a little bit on the macro situation. And then as we go into the second half of the year, in addition to Genoa, we're also ramping Bergamo. So that's on track to launch here in the second quarter and will ramp in the second half of the year. And then as we get towards the end of the year, we also have our GPU ramp of MI300. So with that, we start the ramp in the fourth quarter of our supercomputing wins as well as our early cloud AI wins. So those are all the factors. Of course, we'll have to see how the the year play about how we're positioned from an overall product and road map standpoint for Data Center.”
Below, is where the year-over-year growth rate was confirmed:
“Stacy Rasgon
For my first one, Lisa, can you just like clarify this explicitly for me. So you said double-digit Data Center. Was that a full year statement? Or was that a second half year-over-year statement? Or was that a half-over-half statement for Data Center?
Lisa Su
Yes. Let me be clear. That was a year-over-year statement. So double-digit Data Center growth for the full year of 2023 versus 2022.
Stacy Rasgon
Got it. Which just given what you did in Q1 and sort of are implying for Q2 needs something like 50% year-over-year growth in the second half to get there.So you're endorsing those — you're endorsing that now?
Lisa Su
I am…
Jean Hu
Yes, your math is right.”
Comments on the Client Segment Stabilizing in H2:
Arguably, how the Client segment performs is equally important to how data center performs simply because this segment has been a bit awful for a wide range of companies.
Per the opening remarks:
“Now turning to our client segment. Revenue declined 65% year-over-year to $739 million as we shipped significantly below consumption to reduce downstream inventory. As we stated on our last earnings call, we believe the first quarter was the bottom for our client processor business. We expanded our leadership desktop and notebook processor portfolio significantly in the quarter. In desktops, we launched the industry's fastest gaming processors with our Ryzen 7000 X3D series CPUs that combine our Zen 4 core with industry-leading 3D chiplet packaging technology.”
There were questions in the Q&A with the CEO, Lisa Su, saying the following:
“Lisa Su
Yes. So we've been undershipping sort of consumption in the Client business for about 3 quarters now. And certainly, our goal has been to normalize the inventory in the supply chain so that shipments would be closer to consumption. We expect that, that will happen in the second half of the year, and that's what the comment meant that we believe that there will be improvements in the overall inventory positioning.
And then we also believe that the Client market is stabilizing. So Q1 was the bottom for our business as well as for the overall market. From what we see although it will be a gradual set of improvements, we do see that the overall market should be better in the second half of the year. We like our product portfolio a lot. I'm excited about having AI-enabled on our Ryzen 7000 series. And we have leadership notebook platforms with Dragon Range. Our desktop road map is also quite strong with our new launch of the Verizon 7000 X3D products.
And so I think here in the second quarter, we'll still undership consumption a bit. And by the second half of the year, we should be more normalized between shipments and consumption, and we expect some seasonal improvement into the second half.”
Comments on the MI300 Release:
The MI300 will be released in Q4 and is expected to contribute to revenue by early 2024. I will expand on this in the upcoming deep dive. Here is what was said in the opening remarks:
“Customer interest has increased significantly for our next-generation instinct MI300 GPUs for both AI training and inference of large language models. We made excellent progress achieving key MI300 silicon and software readiness milestones in the quarter, and we're on track to launch MI300 later this year to support the El Capitan exascale supercomputer win at Lawrence Livermore National Laboratory and large cloud AI customers.”
There were many questions about the MI300, one of the more important one being from Ross Seymore who hints toward Nvidia being hard to compete against. Lisa Su’s answer points toward potentially undercutting Nvidia on price and also having strong relationships from EPYC CPUs:
Ross Seymore
And pivoting for my follow-up on the AI side and MI300. I just wanted to know what you would describe as your competitive advantages? Everybody knows that's a market that's exploding right now. There's tons of demand. You guys have all the IP to be able to attack it. But there's a very large incumbent in that space as well.
So when you think about what AMD can bring to the market, whether it's hardware, software, heterogeneity of the products you can bring, et cetera, what do you think is the core competitive advantage that can allow you to penetrate that market successfully?
Lisa Su
Yes. There's a couple of aspects, Ross. And — yes, since we haven't yet announced MI300, all of the specifications will — some of those will come over the coming quarters. MI300 is the first solution that has both the CPU and GPU together, and that has been very positive for the supercomputing market.
I think as it relates to generative AI, and we think we have a very strong value proposition from both a hardware and again, it's a performance per dollar conversation, I think there's a lot of demand in the market. And there's also — I think given our deep customer relationships on the EPYC side, there's actually a lot of synergy between the customer set between the EPYC CPUs and the sort of 300 GPU customers.
Conclusion:
This is the kind of earnings report where you’d have to listen to the call to piece it together. Journalists and others who cover dozens of stocks will rush the conclusion (beat this quarter, lower guidance quarter) but there a lot of depth to AMD right now – perhaps more depth than any other stock we own right now since the product road map is loaded for the second half.
I’m still feeling Zen after this earnings report and also feeling like it’s time to dedicate a new deep dive to the company. It’ll be a bit nostalgic to come back to Xilinx again as I called the company “a heavy hitter AI stock” four years ago. The MI300 GPUs are exciting but Xilinx is also a large piece to AMD’s AI opportunity. Keep an eye out for this deep dive in the next couple weeks as we carefully, patiently, and with an exceptional level of due diligence, build out the world’s very best AI portfolio.
