GSIT - NASDAQ

Ladies and gentlemen, thank you for standing by. Welcome to GSI Technology's Second Quarter Fiscal 2023 Results Conference Call. At this time all participants are in a listen-only mode. Later we will conduct a question-and-answer session. [Operator Instructions] Before we begin today's call, the company has requested that I read the following Safe Harbor statement. The matters discussed in this conference call may include forward-looking statements regarding future events and the future performance of GSI Technology that involve risks and uncertainties that could cause actual results to differ materially from those anticipated. These risks and uncertainties are described in the company's Form 10-K filed with the Securities and Exchange Commission. Additionally, I've also been asked to advise you that this conference call is being recorded today, October 27, 2022, at the request of GSI Technology. Hosting the call today is Lee-Lean Shu, the company's Chairman, President and Chief Executive Officer. With him are Douglas Schirle, Chief Financial Officer; and Didier Lasserre, Vice President of Sales. I would now like to turn the conference over to Mr. Shu. Please go ahead, sir.

Good afternoon, and thank you for joining us to review our second quarter fiscal 2023 financial results. Revenue in the second quarter of fiscal year 2023 grew year-over-year by nearly 15% to $9.0 million, the midpoint of our guidance range. Revenue growth was primarily due to increased sales of high-performance SRAM and the shipment of radiation-hardened SRAMs in the quarter. As a result, our gross margin rose by 900 basis points year-over-year to 62.6%. Higher gross margin, along with flat operating expense year-over-year led to a lower operating loss for the quarter. Net loss for the second quarter narrowed to $3.2 million. As of September 30, 2022, GSI Technology's cash balance, which includes cash and cash equivalents as well as short and long-term investments was $38.9 million. We have a number of items in the file with Gemini-I, but our team is primarily focused right now on developing two near-term markets for the APU. The first is Synthetic Aperture Radar or SAR, image processing acceleration; and the second is fast vector or neural search optimized for big data applications. Let me help you understand why these two markets have been prioritized. At the end of September, we announced that our Synthetic Aperture Radar image processing acceleration system using GSI's APU was approved for use by Elta, a subsidiary of Israeli Aerospace Industries. It is a chance to showcase the technological superiority of our APU and will open future opportunities for GSI in the view of accelerating SAR image formation with new customers. The features that make our SAR image processing acceleration system solution very attractive are: one, scalability; two, low power consumption; and three, portability. First, with respect to scalability. The scalable APU architecture allows expanding to multiple boards on servers for added performance and redundancy without specialized links. The APU platform has the capability to stack servers together and bring near real-time capability to time-consuming, compute-intensive processes. Second, with respect to lower power consumption, GSI has demonstrated using a large area, SAR image process in one second in high-resolution scenarios. That APU used, on average 88% less power than CPU or GPU systems and require significantly fewer servers. And third, with respect to portability, our solution uses one tenth the number of servers than CPU or GPU system, making it small enough to be installed on UAVs and at homes. GSI's SAR can support a broad range of platforms of both from 500 meters quadcopters to 750 kilometers satellite as well as data center deployment. The SAR market is an attractive market where we have existing relationships with prospective targets. Didier will elaborate on the go-to-market strategy for the SAR application in his comments to follow. Our fast vector or neural search product, which I will refer to as our FVS product supports on-prem, hybrid and cloud users. The FVS plugin can be used by customers who are all on-prem and those that may have hybrid storage and a processing in the cloud. For completely cloud-based customers, the Searchium.ai website is live and available for trials. We currently have about 12 users exploring the platform. These users are primarily data scientists evaluating Searchium's capability for their organizations, meaning they lower database and create a front end for the teams to search. The compelling features of the FVS product must be combined with a small hardware footprint and lower operating power consumption compared to traditional large platforms used for search in large complex database. We still have work to do on our end in developing business opportunity for FVS products and scaling the business. The term for the specific searching market we are pursuing is a very large market. Google has estimated more than $300 billion is loss due to search abandonment each year in the U.S. alone. We are providing an accelerator for existing search platforms such as OpenSearch and the Elasticsearch that accelerate the search with lower power and less hardware rather than just adding more traditional computation capability. The Gemini-I plugin expands search capability beyond text and [indiscernible] using a simple Approximate Nearest Neighbor or k-NN vector similarity search. This is Gemini-I's sweet spot in performance and in IT application for the technology. We are ready for production supporting customers on-prem or hybrid solutions. Searchium.ai allow FVS customer to try our product prior to committing their workloads or purchase recovery. We also support full cloud-based customers utilize the OpenSearch database. We have launched Searchium.ai and attract more than 500 visitors and are now working with a dozen opportunities. We will have plugins for large search database. With these three potential customer source to grow this market, on-prem, hybrid and cloud, we are confident that we can see revenue from FVS products in calendar 2023. Last, let me give an update on the compiler stack. We released the [copperhead], the first version of the full stack [indiscernible] compiler for APU co-development for alpha users, and we plan to migrate APU application library over the next few quarters. Currently, copperhead is still in alpha mode and is used exclusively for internal development. We are targeting an end of the year beta release and expect the compiler to be broadly available this calendar year. It is a longer time line than we have originally anticipated, but we are building a fast and efficient tool for use in the operation for [indiscernible] generations. We believe the additional development time will yield better outcome. Now I will hand the call over to Didier, who will discuss our business performance further. Please go ahead, Didier.

Thank you, Lee-Lean. In the second quarter, we shipped Rad-Hard chips to the U.S. prime contractor who use them in prototyping a new program. This was a one-off sale, and we do not have visibility for a follow-on order from this customer at this time. However, if a few of the recent prototypes we've shipped moved to production, it could add incrementally to the fiscal years 2023 and 2024 revenues. As Lean mentioned earlier, we have finalized the SAR POC with IAI/Elta, [indiscernible] we have finalized all the sales and marketing materials needed for customers, including benchmarking, technical papers and an online demo. We are now aggressively launching a campaign to engage with the leading players in the SAR market. It's the early stages of developing this market for Gemini-I, but we have a great solution that is extremely flexible and can support a large variety of use cases. Typically, store applications are very specific but Gemini-I can be using a wide range of altitudes, resolution and image sizes. Also due to the small footprint, our solution can be deployed in small drones or satellites. As a result, we are confident we can successfully build a customer base with this application. Let me switch now to customer and product breakdowns for the second quarter. In the second quarter of fiscal 2023, sales to Nokia were $1.2 million or $13.6 million of net revenues compared to $1.9 million or 23.8% of revenues in the same period a year-ago and $1.3 million or 14.7% of net revenues in the prior quarter. Military defense sales were 22.4% of second quarter shipments compared to 27.4% of shipments in the comparable period a year-ago and 22.3% of shipments in the prior quarter. SigmaQuad sales were 58.1% of second quarter shipments compared to 52.4% in the second quarter of fiscal 2022 and 44.8% in the previous quarter. We remain confident we can fulfill the orders that we have in hand for the upcoming quarters for our SRAM customers and that we have sufficient capacity to meet the demand. I'd like to hand the call over to Doug. Doug, go ahead, please.

Thank you, Didier. We reported a net loss of $3.2 million or $0.13 per diluted share on net revenues of $9 million for the second quarter of fiscal 2023 compared to a net loss of $4.6 million or $0.19 per diluted share on net revenues of $7.8 million for the second quarter of fiscal 2022 and a net loss of $4 million or $0.16 per diluted share and net revenues of $8.9 million for the first quarter of fiscal 2023. Gross margin was 62.6% compared to 53.6% in the prior year period and 60.2% in the preceding first quarter. The changes in gross margin were primarily due to changes in product mix sold for the three periods. Total operating expenses in the second quarter of fiscal 2023 were $8.8 million compared to $8.7 million in the second quarter of fiscal 2022 to $9.3 million in the prior quarter. Research and Development expenses were $6.4 million compared to $5.9 million in the prior year period and $6.6 million from the prior quarter. Selling, general and administrative expenses were $2.4 million in the quarter ended September 30, 2022, compared to $2.8 million in the prior year quarter and $2.7 million in the previous quarter. Second quarter fiscal 2023 operating loss was $3.2 million compared to $4.5 million in the prior year period and $3.9 million in the prior quarter. Second quarter fiscal 2023 net loss included interest and other expense net of $14,000 and a tax provision of $37,000, compared to interest income and other expense net of $8,000 a tax provision of $42,000 for the same period a year-ago. In the preceding first quarter, net loss included interest and other expense of $26,000 and a tax provision of $60,000. Total second quarter pretax stock-based compensation expense was $661,000 compared to $716,000 in the comparable quarter a year-ago and $638,000 in the prior quarter. At September 30, 2022, the company had $38.2 million in cash, cash equivalents and short-term investments and $628,000 in long-term investments compared to $44 million in cash, cash equivalents and short-term investments of $3.3 million and long-term investments on March 31, 2022. Working capital was $42.2 million as of September 30, 2022, versus $45.8 million at March 31, 2022, with no debt. Stockholders' equity as of September 30, 2022, was $58.7 million compared to $64.5 million as of the fiscal year ended March 31, 2022. Given our current cash burn, we remain confident that we have sufficient funds to cover further R&D in the AP platform as we approach breakeven. We're also looking at opportunities to monetize other assets on our balance sheet should additional cash be needed. Given the current global economic environment, our current expectations for the upcoming third quarter are net revenues in the range of $6.3 million to $7.3 million with gross margin of approximately 53% to 55%. Operator, at this point, we will open the call to Q&A.

At this time we will be conduction a question-and-answer session. [Operator Instructions] Our first question comes from Mike Mork with Mork Capital Management. Please proceed with your question.

Hi, [indiscernible]. My question is your revenues are $35 million, $37 million but you're spending $25 million on R&D that seems exceptionally high. And obviously, hopefully, you're going to get some good products out of that. But are you going to keep that at that $25 million? Or is it going to come down, it's going to go up with revenues? And when do you think you will be at a break-even point?

Yes, we are looking into it, and we will talk about it in the next quarter, but I think we will [indiscernible] and but we want to keep the product going. So we are reviewing.

Mike, are you still there?

Yes, I'm here.

Yes. Sorry, the response was that we will be reviewing the R&D costs over this quarter, and it will be – there'll be some kind of an announcement by the next call.

Okay. And how about breakeven? Do you have any guesstimate as one, two years out or what?

We could break even in a couple of years or so. It really all depends on what we see from the Rad-Hard revenue and APU revenues, they both have very high gross margins. And that – once we start shipping will lead to breakeven and profitability.

Okay. Well thank you.

Thanks, Mike.

Our next question comes from [Luke Gabeni]. Please proceed with your question.

Hi. [Indiscernible] the call. Question about the Elta deal. You also got confirmation. I guess, all the testing went well for that. So wondering what the time line is for the initial revenue. And then if you'll have a roadmap for expanding into other SAR clients through the Elta platform?

Great question. So we delivered the initial system, the POC system to them in the summertime. And they're going through their testing and trials now. And so they're looking at possibly ordering two more systems that are slightly different in the first half of calendar 2023. And I'd say slightly different because they have many use cases, which I talked about in my script. As far as other SARs, absolutely. So they're all going to have a slightly different approach. I'm sure, but the meat of what we've done for Elta will be certainly usable for other SAR manufacturers. And as I mentioned, we have everything in line now. We wanted to make sure we had everything available before we put too much hardcore press on them. And so that is done, as I mentioned, all the collateral. We've done all the benchmarking versus CPUs and GPUs. We have all the collateral, all the technical data. And then very recently, we added a demo on our Searchium website that allows customers to actually run our data sets – I shouldn't say our data sets, run a preloaded data set of SAR images, so they can see how it works. And so that's all done now. And we've actually have already started the outreach to these customers and have already gotten responses and are setting up meetings now. So yes, this is going to start happening very quickly.

That's excellent. Do you have any other insight into the SAR market, thinking about – you said there are multiple different SAR applications that are kind of heterogeneous. Wondering about the kind of the timeline of turnover if you know about the legacy systems and when those are due to be out of date and probably looking to upgrade with you or somebody else? More about that maybe the total SAR market with a quick search would think about $3.5 billion maybe $4 billion?

Yes. Yes. So when I mentioned we have a flexible system. So the POC we did can put these – put our solution on a quadcopter. It could put it on a satellite. You can have different image sizes from a kilometer to 10k by 10k images. And so yes, that's very flexible on what we can do. So most of the folks we're going to deal with are going to want one – one is very specific altitude with one specific image size. And so that's what – that's why it's going to be – we're hoping a little simpler with the guys going forward compared to what we did with the POC with Elta, which was a very broad widespread POC.

Okay. That's good that you got that work satisfies that. Do you think you'll be able to – or do you have any kind of rough estimate of what percentage of the SAR market you all could possibly address?

We should have a better feel for it by the next call. As I mentioned, we've done the hardcore outreach very recently, and we have started setting up the meetings. And so it's early right now for me to make that assessment. But by the next call, we should have much more data.

Okay. Kind of remind you of broad – or not a broad question, but kind of a general question about whether Gemini-II will vastly increase the performance of Gemini-I? Or is Gemini-I – it seems like to me is like sufficiently representing the power of Gemini for prospective clients and customers that are testing it now?

Gemini-II will be a drastic – performance drastically improve, okay? So we are looking at maybe an order of magnitude performance improvement. But more important is we – at present, Gemini-I has [FPGA on board] and the Gemini-II [indiscernible]. So from a cost point of view, it's about one quarter of the cost Gemini-II over Gemini-I. So, if you combine performance and the cost, and also the footprint and Gemini-II because of the performance is much lesser cost than the Gemini-I. So if you look at the performance, power, footprint, and the cost, we are talking about more than order magnitude improvement over Gemini-I. And Gemini-I, right now, we are very limited on the AG application. But with Gemini-II, we can heavily involve or go into the AG application. So that's a very exciting prospect of the Gemini-II, and we are very excited about it.

Excellent. Yes. And yes, it's very lower of the higher profit margin and higher performance. And I was making sure that would you all have currently, which is – it seems like the platforms you're setting up you have a software-as-a-service platform. And all this leads to the other question, the [Muves] platform. It seems like you all are partnering with and powering that you all are using the Gemini-I are able to represent this and give perspective clients taste the strength of Gemini so they can extrapolate into what's possible and then potentially start building into those partnerships that you now can look at Elta?

Yes, absolutely.

That was kind of the question and thinking out loud. Okay. Great. And do you have anything to share about the Muves? I saw the video on the website, but I didn't get a clear essence of what all is going on in the background and what your relationship is with it?

Relationship with who?

With Muves. [Indiscernible]

Yes. Okay. So that's, yes, right now, the video you're referring to is the multi-mill video, yes. So we have been doing some work with them. And that was the first demo and the first program or project, I should say, we worked with them. And like you said, it was certainly a very interesting kind of product solution, we'll be able to go with that. And we're looking to expand that to do other types of applications as well. But yes, that is – that started in kind of in the summertime, and we plan to expand among that relationship.

Excellent.

Our next question is from [James Poko]. Please proceed with your question.

Hi. Thanks for taking my questions today. I had a couple of questions regarding the software that Mr. Shu talked about that I believe what was due in July and is pushed out to what date now?

Are you referring to the compiler?

Yes, copperhead. Yes.

Yes. Right now, we plan to early next year, the calendar year, so we can promoted to the beta customers and the general availability. So yes, we are very close to it now.

Okay. Very good. Why didn't we get it done in July? What's the reason why it's now at the – I guess, in calendar year early 2023 now. What happened? Could you explain to us what's going on?

So it actually is being used right now. So it's being used internally. So all of our algorithms that we're developing now or converting are being used on copperhead. So it is being used internally, and we do have a few alpha customers. We just – we just wanted to make sure that there was – it was bug-free and all the documentation was ready before we released it to beyond that group. So the answer is it is here now. It's being used now. It's just not released to the general public yet.

As the first caller said, he said, you guys are burning through a lot of cash. And right now, it's at $39 million. That's not unlimited. You know what I'm saying, especially when we're not generating more cash flow. Could you tell me more about this Rad-Hard prototypes and how they could add incremental sales in 2023 and 2024. How likely is that?

So we've – we have already shipped out product to about – I'm trying to think of the exact number, about eight different entities and a few of the entities have ordered them for multiple programs. And so they're all prototypes. And so obviously, they need to take the parts. Some of them are going to be actually launched. If you recall, we shipped some – our first Rad-Hard was in June of last year. And at the time, we were told that the satellite would be launched at the end of last year, beginning of this year, which would have given us the heritage. And it still has not been launched, and that is obviously completely out of our control. And so the answer is these – once they do the proof-of-concept and the prototyping that it's real, then that's when they go into the production phase. And we don't have that visibility yet on those programs. But certainly, we anticipate that some of these programs will certainly go into production.

When? When?

That's a good question. Again, it's...

Well, we need income. We need more sales. And we're burning through cash. We have delays in the development of the APU, which actually sounds like it's unique technology based on what you said, and I read the marketing thesis online about the SAR marketing piece, it looks tremendous. But we have a share price that's trading at or below liquidation value. Stocks don't get at those levels for no reason. Usually, it's when there's poor execution by the strategy by the management team. So I guess my point is we're trading at liquidation values. We're not getting any valuation from the APU or even the legacy business, it looks like it, too. What can we do about that to get investors to recognize like the benefits of this APU and the technology we have maybe get some excitement there because it doesn't seem to be – we don't seem to be getting any traction in that area.

Yes, it's a new process coming to the market. So it does take a little time from the design to the product to the revenue generation. I know it's been taken couple of years. But that is the period, the new device we have to go through and do – we do have some learning curve going through. And hopefully...

We need to speed that effort up. One other thing I would like to make a comment about January 20, 2022, the largest institutional investor in the company made some recommendations to the company that I don't believe any of them have been implemented, at least one, and his recommendation was he wrote a letter to the Board, recommending a need for a new leadership who's better qualified to execute and lead the company. Right now, we haven't taken any action on his recommendations, and we have a stock trading at like basically liquidation value. I highly suggest to the Board of Directors and Mr. Shu consider that. I think the company could use a little boost of some sort to these very exciting technology, especially APU because right now, I don't see it happening. And we are running out of resources. We need to protect that resource so we can get to the next level. But thank you for your time.

Thanks, James.

[Operator Instructions] Our next question comes from George Gaspar. Please proceed with your question.

Thank you. Could you zero in on this commentary on the expectation of lower revenue stream in the existing quarter that you're in now? Can you tell us where is the lower revenue coming from? Where are you going to have lower revenue? So what product line within there?

So the first one was we shipped a fair amount of revenue associated with that Rad-Hard prototype that I talked about. And as we mentioned, that was a one-off sale until they can go into production. So that was certainly some – we had also two programs that did a buildup of inventory from two separate customers that were significant customers for us and the backlog isn't there for this quarter. So we've talked to them and certainly the products are healthy. They just have gotten into an inventory position. And so that's really the three main areas, and they were significant.

Well, with that being the case, there – you can't see or be looking for very much of any kind of actual net revenue in this quarter from what you're trying to apply into the Gemini market right now, correct?

Okay. I'll make sure I have the question correct. So you're asking if this quarter, if the revenue has any APU in it?

Correct. Correct.

So it has minimal revenue. So we do have two customers that we will be shipping a system to and that's for them to do their trials. And so it will have minimal revenue. It's really – they're demo systems for them to evaluate for future production purposes.

I see. So there's just nominal. And then, of course, the outlook is going forward beyond this quarter, that we're probably looking at more into what like the second calendar quarter of next year before something can really develop and on the Gemini-I in terms of revenue streams, solid revenue stream.

We don't have a time line exactly because, again, we've had – we've had systems in multiple companies hands and they're doing their evaluations. It's – they haven't given us an indication on when that would progress beyond that. So to say, second quarter next year, I don't know. It could be – it may take a little longer.

I see. Okay. Can you answer this question, the relative differential between Gemini-I and II, now that you're working forward on Gemini-II, do you think that this might deter your potential customers from jumping in and testing and going for actual purchases of Gemini-I until Gemini-II is out there where they could compare it with Gemini-I? Or can you just explain that a little bit?

Sure. So – and this is going to expand what Lee-Lean said. Gemini-II is actually going to really help us more on the Edge. I mean, can it do everything Gemini-I can? Yes. But what we're using Gemini-I for now is sufficient. I mean it's – the performance is exceptional. What it lacked Gemini-I was – and again, going back to Lee-Lean's comment because of the FPGA that was on the LiDAboard with Gemini-I. The cost of the solution was higher. The power because of that FPGA was higher, the form factor was higher. With Gemini-II, we don't need that FPGA. So now the Edge application, which was a very challenging and if not impossible, Edge solution with our part. And so now with Gemini-II, we'll be able to address those edge applications that Gemini-I just was not a good fit for. So it's really we're trying to address a different market with the Gemini-II. I should say, a different application really. It's the Edge application.

Okay. All right. And then lastly, as far as the total cost structure associated with stock that's issued to staff. Can you compare the last quarter with the previous quarter? And what – how many shares are we talking about that are being cased into staff on a quarterly basis at this point in time. And considering the fact that the price is down so far now, how do you going to handle that?

Well, in the most recent quarter, a quarter ago, I think it was only several 100,000 shares. It wasn't that much. And for the most part, that is somewhat typical – if you look at the 10-Q, you can see the activity for the most recent quarter, the June quarter, and you can also look to the 10-K, it's got all the activity there. And that should be somewhat reasonable estimation of what the future might behold.

Right. Well, specifically, with the price and where it is now versus what the average price is for the 700,000 shares that you're talking about that were issued in the last quarter, does this suggest that your outlay of shares will be a lot more going forward based on the fact that the price is lower?

No. We don't issue the shares based on the price. We have a structure within the company based on person seniority and position and so on shares we typically ran to that kind of position. The price has no bearing on the number of shares.

Gotcha. Okay. All right. Thanks for that explanation. Thank you.

Our next question comes from Chris Mackay. Please proceed with your question.

Yes. Thanks for taking my call. In light of the cash balance, are you expecting a share dilution possibly within the next 12 months? Thanks.

Are we anticipating issuing more shares? Not at this time.

Okay. Thanks.

Ladies and gentlemen, we've reached the end of the question-and-answer session. And I would now like to turn the call back over to Mr. Shu for closing remarks.

Thank you all for joining us. We look forward to speaking with you again when we report our third quarter fiscal 2023 results. Thank you.