Theo
3週前
My synopsis of the CC 10/29/24
Part three
{MY comments are in BLUE}
Q&A
Richard Shannon-
1. A transformative situation. Maybe you could just give us a little bit more detail on exactly what this means. Any idea on technology area? Sounds like it could be a number of months before it closes.
A. I would say this is a large customer with very significant revenue potential that could really, if we get into production with these guys, will provide a really interesting financial position for us- if we can work with a company that is very large and could deliver high revenue, then that will certainly be very, very compelling for us. We feel great about business there, and we've been working with them for some time… we've been working together for quite a while, and we have a meeting of the mind that it makes sense. This is one of those proposals that we've had outstanding before, and now we've gotten into active negotiations. I can't really say much about the segment they're in, or the timing. But I can say that it's something we're quite excited about.
{Scott certainly seems incredibly confident}
2. You've got a very good funnel with half of the top 20 largest semiconductor companies in the world in the funnel. But I think everyone would like to see you get across the finish line with a qualification… Maybe if you can talk about to the degree to which you're seeing progress with others that are far down that funnel… give any detail as to technology area or areas you think could be among the first ones:
A. The ones in Phase 4, obviously, they have the technology installed in their fab, and they've run wafers and are making good progress in those segments.The customers in Phase 3 are the ones where we're in active experiments with many of them. Some of the legacy technologies that we work with, they take a bit longer to get to production. They take longer to decide to change their process node. And then once they're ready to make that change, they go through a very long process of doing that. I would say ST is one that falls into this category. We believe that some of the more advanced node customers that we've been talking about, like, in gate-all-around and memory, will move faster, but they'll require huge amounts of resources. We know that in those companies, they have massive teams of engineers working on these new nodes. And when we're working with them, they ask us for a lot of data, and they're constantly pushing us to do more and more experiments for them and with them. But we do believe -- because there's so many people on their side and there's such a big push to get to production that they'll be faster time to market. So, that gives you a little bit of insight into the work that we have going with the many different customers that we have underway right now.
{So Scott still thinks GAA will be "first to market"}
3. I guess I want to follow up on RF-SOI… an advancement in the industry with going toward thinner wafers… that was one of the significant delay factors when we've been hearing about RF-SOI potential for a number of years now… Maybe give us an update there of how much that has pushed the ball forward
A. Earlier in the year, we announced an agreement with Soitec to the point where they would make and sell engineering wafers to our customers so they could run more tests to bring that type of technology to market. And we've been just working with that on our customers that are working in this area. And unfortunately, no announcements yet
4. I guess this is probably a question more on the logic side, but is your work ongoing with both foundries and fabless players alike?
A. On the advanced node, no. For the most part, we're working only with people who manufacture the gate-all-around And so, we'd be targeting primarily the folks who have the factories that are going to run these chips.
5.Then my last question, just hitting on JDA 1 specifically- Maybe if you could touch on the specifics there, please.
A. I would say, it's certainly typical that you bring a customer a lot of data showing, hey, we can really do this. And then they give you a request for 10 more pieces of data to validate all of their doubts. And so, that's the stage that I'd say we're in right now
6. Questions taken from retail - is the recent announcement about Sandia Labs related to the CHIPS Act application submitted in July?
A. "No, it's not. It's independent of that, which is great because I think it opens up another avenue for us to demonstrate the GaN technology.
With the CHIPS Act opportunities, we did submit one that was relevant to GaN, but also to other applications of MST in compound semiconductors. So, basically on Sandia, we had proven out the physical benefits of MST in building a GaN on silicon wafer in terms of better quality because of all the stress that's created by building GaN on silicon and something that we could improve there. But at Texas State, we had no ability to actually test electrically how a device would perform when built on that. What we found at Sandia was they had a window under a rapid access program to get us in and start testing how, when you built GaN devices on those substrates that we had worked on with Texas State, to show that they would actually perform better in addition to the physical qualities being improved.
So, the nice thing about that is the Sandia work is at no cost to us, but what we prove there should make us more attractive for both CHIPS Act funding, but also for the several commercial customers that we've gone out and spoken to about our GaN offering. And the last thing I can tell you is I have seen many opportunities for GaN and other compound semiconductors where there are funding opportunities there. So, that's something we'll continue to pursue. But they are separate.".
7. Are all the proposals we have outstanding are currently in the pipeline in, for example, Phases 1 through 3.
A. "yes, all of our current proposals are with customers that are in our pipeline. They're not all in Phase 3, some are in Phase 1 But they are all people that we've been doing work with for some time."
8. Frank, there are questions about plans for funding. Do you want to address the outlook for funding and cash needs?
{This is the question that pertains to the future funding of Atomera I mentioned earlier in the Financial summary}
A. "We ended the quarter with a lower cash balance than we did at the end of June. And obviously, we didn't want to be selling any more stock under the ATM program than we had to, given that the stock in the last number of months was at pretty close to historically low levels. We always have to balance not diluting our shareholders with the fact that we need to maintain a minimum cash balance of greater than 12 months of what we plan to spend in the future.That's just a requirement under GAAP accounting. So, that's always going to be the balance that we maintain. And to the extent that we can be more conservative when the stock price appears to be relatively low, we will do that.".
{Based on the previous $3mil per Q burn, that would mean $12 mil/year. They have $17mil in cash at the present so they "shouldn't" have to be in a rush to tap the ATM before they end of the year imo.}
9. Richard asked: To what degree is the work that you're doing in this voltage range specific to 48-volt that's related to servers, versus more broadly across the space here?
A. The 48-volt stuff is very specifically tied to this expanding opportunity in data centers… it's the same type of transistor structure that would be used for other 48-volt power requirements. But I think the big opportunity is located there at the data center right now. I have to say, I'm quite excited about it. The results that we've gotten from testing show really good efficiency improvements that I think will be very attractive to people, but we're just starting to talk to customers about it today.
{Obviously very early in the game but let's hope the efficiency improvements really are that good or impressive ...}
10. If you're trying to insert yourself into legacy nodes here, the time frame to get to production or to get to a license and moving forward seem to be more delayed than with newer nodes here. I guess, is there anything you can do to help accelerate that in any way? Or is this just a relic of working with companies with legacy nodes?
A. We do believe when ST hits the market, that's going to make a big difference. And also, some of the proposals that we have outstanding are with companies that would be entering into that space. So, I think we'll get more even before ST goes to production. That's my belief. if you think back about the history of ST, we started working with them in 2017 or 2018. We did some experiments. By the end of 2019, we really had shown excellent results with them. And between COVID and then building a new factory, they didn't really start work on the new process node until 2022. So, it's not that they take a long time once they decide to go to production. I think that part is maybe slightly slower, but similar to another process node going to production. But the decision to make a change to an old legacy node that's been running in high volume for some time, that doesn't come up every year. And so, sometimes we have to wait for that to happen.
11. Generally, how is the relationship with ST going, and are there more comments you could add to give us a sense of the progress?
A. "ST is excellent company to work with obviously, we're working with some of the best technical experts in the field here. And I think we really enjoy the back and forth to be able to constantly be improving the products. Things have been going so well with ST that we've also executed a land-and-expand type of strategy where we've already started working with some other groups inside ST as well. And so, hopefully, we'll be able to expand our business into them to multiple business units. And it's possible we could do that even before we get to volume production with the first product. So, things are going very well with them."
{Always nice to get further clarification and confirmation. As I've stated numerous times over the years, you keep showing me "progress", I'll stick with it}
12. And just one final question. Can you provide some comments on the gate-all-around silicon results that were presented at PRiME 2024?
Nice little explanation for those of us a little slow on the uptake when it comes to semiconductor technology like me.
A. "OK, let me explain this. In a gate-all-around source and drain, it's very high-doped phosphorus source or drain, and then it's going to go into a channel where they don't want any doping at all. They'd like that to be almost dopant-free. Since the structures are so small, it's difficult to keep that gate, which is highly phosphorus-doped, from diffusing into the channel. And if it diffuses into the channel enough, it can actually cause a short circuit between the source and the drain, or at least such low performance that the device won't yield. So, that will be a low-yield product. We believe that by putting MST at that interface between the source and the channel that we will help prevent phosphorus from flowing into the channel side and either shorting or having lower performance. So, number one, it will lead to higher performance. But even more importantly, it will help, we believe, the gate-all-around manufacturers to yield a higher number of transistors when they're in manufacturing. So, that's one of several applications that we have in gate-all-around that we're promoting to customers."
Theo
3週前
My synopsis of the CC 10/29/24
Part one
{MY comments are in BLUE}
Scott's opening statement points:
1. The last three months for Atomera have been the busiest and most positive in my memory. {If not hyperbole, It's a very nice development}
2. Atomera has been supporting ST in the development of their new process for over a year now.
a. The first is on MSTcad simulations to optimize silicon performance through integration of MST. This work has been underway for over a year and the correlation between modeling and hardware has been confirmed with real silicon multiple times.
b. Second area in which we are cooperating with ST is manufacturability by enhancing throughput of MST deposition on ST's factory production tools. I want to emphasize that this project is in great shape. Indeed, ST has authorized me to say that development is going well and that we are still on track to go to production as soon as development and qualifying are complete. {Two things strike me here- 1, ST & Atom ARE working on a "manufacturing" process. That, to me, suggests they really are advancing imo and 2., ST finally acknowledges, and Scott confirms, to allow him to give an update such as it is. Minor perhaps but still implies that ST is at least aware of Atom's current shareholder frustrations}
c. At that time, we expect royalties that will be compelling high gross margin revenue for Atomera. {We ALL are}
d. ST is not the only company we've been working with in this segment- it's interesting to many different players for a variety of applications 7 – 48 volts {"Many different players"?? Sweet!}
3. We are currently in discussions with multiple customers on proposals to either license the technology outright or to enter into a JDA that will ultimately lead to production. {"Currently" and "multiple customers" are the operative words here imo- Another good sign}
One part of this market is particularly hot right now and provides a compelling opportunity for Atomera.
1. Traditionally, server racks and data centers have been fed by a 12-volt power supply, but that standard is currently changing to 48 volts.
2. Atomera has finalized a 48-volt version of our SPX technology with dramatic performance improvements specifically to address this segment.
3. Next, we have the advanced node work for gate-all-around transistors at the leading edge, which represents about $150 billion market in 2023.
a. As device architecture shifts from FinFET to gate-all-around, the number of epi steps incorporated into the process flow is projected to more than triple. This trend makes it significantly easier to add MST to the primary epi steps… As epitaxy becomes a more critical component of device architecture, it opens the door to our MST technology being more easily slotted into the production flow. {"epitaxy". Pay attention to this in future PR's}
4. Today we are working with multiple customers in the advanced node area. The memory segment, valued at over $110 billion in 2023, has many characteristics in common with the advanced node segment {"Multiple customers"- always nice to hear, just need to convert them to paying customers}
a. Again, similar to advanced logic platforms, the memory platforms are beginning to introduce epitaxy into the device structure, MST becomes an incremental cost adder to the epi steps while delivering substantial device performance and cost benefits. {"epitaxy" again...}
b. And the opportunity here is huge at approximately 20% of the entire semiconductor industry, with very high volumes and long technology cycles. Again, we are working with multiple customers in this area. {"epitaxy" is the keyword here- To understand the importance of it in relation to MST, review the article available on Atomera's website Atomera/Sandia project to address GaN/Si challenges}
5. In each of these segments, we've made proposals to customers which we are optimistic will convert into JDAs or license agreements in the near future. Indeed, we are in active negotiations today on an agreement that we believe will be transformative for the company, and we are hopeful that we will be able to announce it within the next few months. {"Transformative" There's been a lot of buzz concerning that word Scott chose- I have to admit I'm curious (and admittedly a little stoked) as to what that infers...}
6. GaN on silicon will be the high-volume, low-cost substrate of the future for GaN devices, and if this technology works as expected, MST will be valuable to anyone who manufactures with it. Those wafers are experiencing stress that lead to lower yields and higher defects, which MST may help to alleviate. The rapidly growing GaN market is particularly interesting to us because we believe the technology can be converted to revenue faster than our traditional business, and we hope the work with Sandia will help speed up our time to market. {HINT: "epitaxy"}
a. We are engaged with a lot of very large customers, and I sense that the need for our technology is stronger than ever. {"A lot of very large customers"?? Gotta love that }
7. JDA 1 has recently requested additional data from us to validate MST's effectiveness in a specific application, and we are already planning wafer runs to demonstrate how MST can address their requirements.
8. Discussions with JDA 2 about a company wide license spanning multiple technologies and nodes is ongoing, although this negotiation is moving more slowly than we would like. {This one sounds more like a situation that will "domino" once someone else makes the first move}
9. Our fabless licensee has wafers coming out imminently. And if those results are strong, we believe it will lead to a production transition program with both them and their foundry partner. {Always "if". It would be nice if we had more clarity on what the time frame is with the word " imminently"...😒}
Scott's Summary:
" Beyond these engagements, we have proposals outstanding with several others in addition to the large transformative customer we are negotiating with today. {There's that word "transformative" again 🧐} Right now, without exaggeration, I believe the Atomera team is more excited about our prospects than I've ever seen. {Much like his opening statement "busiest and most positive" in memory- gotta be something to that...} The ST engagement, which is an enormous business with a long projected life, is moving nicely toward production, which ST has confirmed. Our potential in the outlined segments is even larger, {Larger than the ST deal??? 😲}} and I feel we are very close to announcing deals which will cement that position. The work we're doing in gate-all-around and memory is tied directly to the biggest driver of the semiconductor industry today, the rollout of AI infrastructure. And our GaN work is not only aligned with a major industry push, but we also believe it can be executed with faster time to revenue than our other segments."
_______________________________________________________________________________________________________________________________
Theo
4月前
Con't...
JDA1
We continue working with JDA1 to move into a more definitive production program, but have not yet achieved this result.
HMmmmmmmm....doesn't sound promising to me...
JDA2
JDA2, however, is in active negotiations with us on a license, now that we have results from wafer runs completed in the last quarter. We hope to make this a comprehensive company agreement covering multiple technologies.
THAT sounds much much better...
Admission of failure
"Progress in converting these engagements to licenses on the path to production has not met our expectations. We know that our ability to develop technology and to build customer interest has been excellent, but we need to do better at closing deals."
You'll get no argument from me on that point...
"Recently, we've taken a hard look at our sales and marketing efforts to find ways to improve. As a result of this process, we decided to make changes that start with bringing on a new Head of Business Development and Marketing, Shawn Thomas"
At least you've made a move rather than simply sitting on your as............laurels. 😒
MST-SPX
This is a relatively new direction and refers to power transistors. (5V to 48V). Interesting and good read.
"I'd like to highlight one of our new technologies called MST-SPX targeting 5 to 48 volt transistors. On Friday, we released a white paper detailing what we've accomplished, but allow me to give a brief summary. We have found that combining MST with the advanced structures used in high-power transistors, allows us to achieve better performance than has been published by any other semiconductor maker. As seen in the chart on slide 5, where lower is better. This is pretty amazing stuff.
We've also calculated that it makes financial sense for our customer to implement MST at less than a 3% improvement in RSP. But what we're seeing and what we're showing on this chart show improvements between 15% and 30%, confirming that our technology offers best-in-class performance and compelling economics for next-generation powered chips.
Let's turn that into revenue!
Validation of Atomera's work and focus:
"At SEMICON West earlier this month, Dave Thompson, Intel's VP of process technology and research gave a talk on the solutions the industry will need in the future for evolving transistor architectures. Virtually, all the examples he brought up, including source drain resistance, channel mobility, dopant control, gate dielectric issues, and even gallium nitride were things Atomera is working on. His message was clear, the industry sees increasing technical challenges moving beyond three nanometers and finding solutions will require a strong cooperative effort by ecosystem partners across the industry.
This is a type of statement that entirely validates the direction we've been taking at Atomera and shows the importance and potentialof the work we're doing today. For this reason, we are very optimistic about the prospects of our company. Our ST engagement is progressing nicely and should form the base of our revenue starting next year. And each of the other segments I've outlined can grow on top of that base."
CHiPS Act
"This past quarter, we also submitted our first of hopefully many proposals under the Chips and Sciences Act. Atomera's potential is being recognized across the industry, and it's only a matter of time before several of these potential opportunities turn into more business prospects for Atomera. We're working very hard to make that happen."
Finally!
Con't...
Theo
4月前
Last CC con't...
DRAM- "...Atomera is working at various stages of adoption with more than one of the major memory manufacturers...., these customers are laser focused on cost, which historically has made them late adopters of new materials that could affect yield or increased costs including via royalties...
...Once adopted though, new materials technologies have long legs and wafer volumes are the highest in the industry. We're excited about the potential for our technology in this space..."
Scott does not exude as much confidence in the DRAM sector at the moment compared to the others imo. Doesn't mean it won't pan out, just that I doubt this sector is advancing as quickly as others. Again, that's just me.
(GAA (Gate All Around) "...advanced node market is one of the largest by revenue and the smallest by device count approximately $150 billion in 2023, more than 28% of the overall semiconductor industry. Atomera is working with the major gate-all-around manufacturers on solutions for the leading nodes down to three nanometers. But we are also exploring development at the nodes leading from two nanometers down into the angstrom range.
The revenue potential for Atomera might be the highest in this market segment because the cost of leading logic chips is higher than other semiconductor products. Our MST technology could add significant value to these devices that are at the heart of the AI revolution, driving higher performance, lower power consumption, and better yield. On our website, we have white papers describing exactly how MST provides these benefits...
... That said, we're not sure when we might be able to announce something specific in the gate-all-around sector, because the development programs are so large and the timeframes are so variable, but the interest in work is real and the upside for us is significant
I get the impression from Scott's dialogue here that there's a real push by Atomera to get into this market. Indeed, he belabors the point of Atomera's white papers focused on this market segment. Great! Rock and roll dude- git 'er done.
GaN (gallium nitride technology) I'm intrigued with the prospects in this segment.
"It might initially appear that GaN is a departure from our core business, but it is not. MST GaN would be applied in the manufacturingprocess in the same manner as the other applications we've already covered to improve the yield of a myriad of GaN related semiconductor products...
...Interest by potential customers and partners has been surprisingly strong. Indeed, we are already in discussions with four potential customers and partners...
...our R&D team is still working to turn this into a full product release. We believe with this high level of customer interest, early revenues from MST GaN might be possible in the near term, even possibly later this year.
Con't...