Last week, Atomera announced its first joint development agreement. In this blog update, we revisit the company that was first covered in March of 2020 in this PDF.
Atomera offers Mears Silicon Technology (MST). The team at Atomera has been developing this technology since 2001. The technology works on an atomic level to insert thin layers of non-semiconductor material to increase the “concentration of dopants.”
Atomera has developed MST wafers with the goal of replacing current specialty wafers by keeping the cost the same but delivering an increase in performance. In more simple terms, the goal of MST is to reduce the effects of Moore’s Law by reducing chip size, delivering higher electron-mobility and increasing the performance of the transistor. According to third-party PMIC and per Atomera’s website, MST can result in 16-21% reduced chip size.
The Joint Development Agreement (JDA) is with a “leading semiconductor provider” and includes “a manufacturing license allowing the customer to fabricate semiconductor wafers incorporating MST for use in their products.” Ideally, we would have been provided a name. In absence of this, I mused on the forum earlier today that perhaps it’s 5G related due to the CEO having quite a bit of mobile on his resume with VP and GM positions in the mobile division of Broadcom and also from Altera. However, until the company announces this, we won’t have any way of knowing.
Theoretically, if one company uses MST then others will follow because MST provides a competitive advantage by reducing chip size and enhancing transistor capabilities. MST also helps memory/DRAM designers by reducing the size of the chip without moving to a new technology node. In the previous report that I published, I outlined that using MST/Atomera would cost around $40 to $60 million total while a foundry for a new node can cost billions of dollars – so MST saves quite a bit here.
In the past, I’ve called Atomera a Hail Mary play. I’d say we are now on the 1-yard line.
Atomera has always been an all-or-nothing proposition as the company is essentially pre-revenue. This is a company where we saw there could be an important need for the product/invention as the semis race heats up. You’re seeing major evidence of the semi race right now and even the smallest advantage can make an impact between competitors.
We attempted a position around March but did not remain in our position because we wanted to allocate our money elsewhere until the first licensing agreement was announced. We like where the company is at now as far as risk/reward.
Atomera is pre-revenue. The company reported only $62,000 in the first nine month of 2020 compared to $395,00 for the same period last year.
The company loses roughly $12 million per year so it’s not surprising the company raised cash recently with a $25 million offering on January 5th, 2021. The company also sold shares of 845,730 for net proceeds of $8.6 million.
Net cash used in operating activities for the nine months ending September 30th was $9.1 million. In the most recent quarter, the net loss was $3.6 million or EPS of ($0.19).
There is only one analyst covering the company with estimates of EPS ($0.62) for 2021 and revenue of $130,000 for 2020 compared to estimates of $3.5 million for 2021.
Revenue is generated from integrated licensing agreements. Customers pay a licensing fee to use MST technology in the manufacturing of silicon wafers. Royalties are paid for each silicon wafer or device that incorporates the MST technology.
When we say an all-or-nothing proposition, what we mean is that Atomera can become profitable off the royalties from one wafer fab company and there are 370 wafer fabs worldwide. The annual revenue at 50% ramp from an average sized fabrication company would be $6.7 million in revenue. The annual revenue at 50% ramp from a larger foundry would be $29 million in revenue at a 2% royalty rate.
Please note, as discussed in an investor’s call in December, the royalty ranges from 1-3% depending on the size of the fab. The larger the fab, the lower the royalty fee. According to one analyst who covers the stock closely, once adopted, the technology will be used for a minimum of 7 years.
The company provided the following in the recent Investors Presentation last month:
If the company is truly “working with 50% of the world’s top semiconductor makers” and the first licensing agreement was announced, then perhaps the rest will line-up nicely (and quickly for our sake).
Target customers (i.e., not current customers but more like addressable market) include Samsung, Micron, Intel, Texas Instruments, NXP, SK Hynix, Qualcomm, Marvell, Broadcom, Skyworks, Qorvo, Taiwan Semiconductors, Applied Materials and more. You can view the full list here on page 3.
The total addressable market outlined here on pages 9-10 is between $6 to $7 billion in royalty fees primarily driven by FinFET and Advanced Nodes ($6 billion), RF SOI ($50 million) and 5V Analog ($660 million).
In a previous Investors Presentation, the company discussed having licenses with STMicroelectronics and Ahashi.
Prior to the announcement, the CEO was cautious to say when a manufacturing license would be issued. From what I’ve seen from Atomera, they’ve always been very diligent to distinguish between the various phases and potential setbacks that testing and integration can cause. The ultimate goal, whether in conversations such as the one below, or in the Investor Presentations, is the licensing as this is what creates revenue.
This is from a conference in September:
“When do we believe it will happen? We have not typically predicted when we will get to commercial production. One of the reasons is when customers work with us, it takes 9 months for us to do a single evaluation run with them. Maybe even longer. Um, and in the past, we’ve experienced a lot of times we have a good evaluation run with a customer, they get a good result and they want to do another one to try to make the result even better. And so that’s how these types of technologies are brought to market. So if we believe that we’re gonna reach the end in 3 months, we may find out that at the end of that we have another 9 months in front of it. We avoid predicting that but we are in the late stages of evaluation and integration with many different companies, and we’re hopeful that we’ll reach a manufacturing license with them relatively quickly.”We avoid predicting that but we are in the late stages of evaluation and integration with many different companies, and we’re hopeful that we’ll reach a manufacturing license with them relatively quickly.”
The CEO grew a $1 billion-plus division at Broadcom and also worked as an SVP and GM at Altera – which was a competitor to Xilinx before being bought out by Intel. He comes specifically from a mobile background. The CTO has been inventing and working on patents for broadband networks for 30 years and is who MST is named after. As I said last March, I don’t see any flags here and I think management has been very transparent leading up to this point.
I look forward to hearing what company the manufacturing licensing agreement is with and how soon the royalty fees will affect the top line.
Please note, we are very early to Atomera and there will always be time to consider the stock later down the line when it carries less risk. That is the beauty of small caps and tracking them early. Anytime there are debates about companies that are higher risk, the answer is almost always “revisit later” for anyone who is not comfortable with the risk profile.
We like the JDA and licensing announcement and we are following a plan we set in May for this company – however, Knox is critical here (more so, than say, a position in Microsoft). Knox is very good at stops and tracking technicals so we are combining a few of Knox’s techniques right now.