Ladies and gentlemen, thank you for standing by. Welcome to the GSI Technology's Third Quarter Fiscal 2020 Results Conference Call. [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 involves 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 have also been asked to advise you that this conference call is being recorded today, January 30, 2020, 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, everyone, and thank you for joining us to review our third quarter 2020 financial results. In 2019, we made significant progress on our increased associated processing unit, Gemini-1. The momentum has continued into 2020.

In March, we anticipate completing the benchmarking of our Gemini-1 chip against industry-standard benchmark for similarity first using FAISS and ANN industry standards with database up to a billion items.

We expect to publish our results in summer of 2020, followed by the qualification for Gemini-1, which we anticipate to be completed in the second half of 2020. Our AI team is finalizing the software library and algorithm for Gemini-1 in preparation for customer workshop.

Gemini architecture is based on the memory sensory design that removed the limitation of cellular processes. Gemini combines core capabilities in memory chip design with our AI software expertise. As a result, massive parallel data processing, compute and search all take place directly in the memory array.

This architecture has increased data processing fees and lowered operating costs by orders of magnitude with a scalable architecture. The last feature scalability, distinguished architecture of our Gemini product. Another distinctive feature is the product for 1 interface.

They allow multiple levels of custom programming from assembly code up to tens of code of Python. We anticipated 2020 to be the year of design win for Gemini-1 for similarity search applications.

The application we are targeting, including signal and opt object identification and the classification, facial recognition and drug discovery for new drug development. Once we have completed the Gemini-1 one qualification, we will also talk to e-commerce customers for future service.

I appreciate our AI team's hard work to bring Gemini-1 to market, and I am pleased with the consistent performance from our legacy business team has allowed us to fund Gemini's research and the development. GSI is bringing a new and novel approach to AI with our in-memory processing technology, which is patent-protected.

And that's why we anticipate will ultimately be a game-changer for speed, power and scale in AI. Now I'll turn the call over to Didier, who will discuss our business performance in further detail. Please go ahead, Didier..

Thank you, Lee-Lean. In addition to all the progress we've made on Gemini-1, we have completed our internal qual qualification on our radiation-hardened SRAM to Class Q and Class V level to meet the requirements of aerospace and defense customers.

GSI's radiation-hardened solution leverages our proven technology and offers up to 4x the memory size with the same footprint and power as the current memory solutions. This means legacy RadHard chips rarely can be switched out for GSI's higher-density solutions at the 40-nanometer technology node.

We have identified several prospective opportunities for both our radiation-hardened and radiation-tolerant chips. As a result, we may see some sales in calendar 2020 in those product categories. We still see that market TAM of $15 million to $20 million for these products.

Last quarter, we also had the opportunity to radiation test the Gemini-1 product and the results were very promising. Gemini-1 will allow data processing on a much greater level on satellites, expanding the markets for Gemini products to include space and aerospace.

Looking at our sales breakdown in the third quarter of fiscal 2020, sales to Nokia were $2.6 million or 26.3% of revenue compared to $6.6 million or 45.2% of revenue a year ago and $5.3 million or 45.2% of revenue in the prior quarter.

Nokia sales have been slower in the most recent quarter, and we expect a similar level in the fourth quarter of fiscal 2020. Looking out to the remainder of calendar 2020, the forecast is expected to improve and return to levels we saw earlier in this fiscal year.

Military and defense sales were 18.9% of third quarter shipments compared to 16.7% a year ago and 23.4% in the prior quarter. SigmaQuad continues to be our best-performing product category, with sales of 62% of third quarter shipments compared to 69.7% last year and 63.5% in the preceding quarter. I'd like to hand the call over to Doug now.

Please, Doug, go ahead..

Thank you, Didier.

For the third quarter of fiscal 2020, we reported net loss of $4.6 million or $0.20 per diluted share, on net revenues of $10 million compared to net income of $2.3 million or $0.10 per diluted share, on net revenues of $14.7 million for the third quarter of fiscal 2019 and a net loss of $1.8 million or $0.08 per diluted share on net revenues of $11.7 million in the second quarter of fiscal 2020.

Gross margin was 60.2% compared to 68.3% in the prior year period and 55.9% in the prior quarter. The changes in gross margin in each of these periods is primarily due to mix of products sold in each quarter.

Total operating expenses in the third quarter of fiscal 2020 were $10.8 million compared to $7.8 million in the third quarter of fiscal 2019, and $8.5 million in the preceding quarter. Research and development expenses were $8.2 million compared to $5.2 million in the prior year period and $5.8 million in the prior quarter.

The increase in R&D expenses reflects a charge of $2.7 million during the third quarter of fiscal 2020 for purchased intellectual property that will be incorporated into the next-generation of our Gemini chips.

Selling, general and administrative expenses were $2.6 million in the quarter ended December 31, 2019, compared to $2.6 million in the prior year quarer, and down sequentially from $2.7 million in the previous quarter.

Third quarter fiscal 2020 operating loss was $4.7 million compared to operating income of $2.2 million in the year ago period and an operating loss of $1.9 million in the prior quarter.

The third quarter fiscal 2020 net loss included interest and other income of $207,000 and a tax provision of $84,000 compared to $96,000 in interest and other income and a tax provision of $70,000 for the same period a year ago. In the preceding second quarter, net loss included interest and other income of $210,000 and a tax provision of $55,000.

Total third quarter fiscal 2020 pretax stock-based compensation expense was $629,000 compared to $592,000 in the comparable quarter a year ago, and $642,000 from the prior quarter.

At December 31, 2019, the company had $68.3 million in cash, cash equivalents and short-term investments and $5.3 million in long-term investments compared to $61.8 million in cash, cash equivalents and short-term investments of $9 million in long-term investments for the fiscal year ended March 31, 2019.

For the upcoming third quarter of fiscal year 2020, our current expectations are net revenues in a range of $8.5 million to $9.5 million with gross margin of approximately 50% to 52%. Operator, at this point, we'll open the call to Q&A..

[Operator Instructions]. And we'll take our first question from Ari Shusterman with Needham..

This is Ari taking the question for Raji Gill.

So first one to start off by talking about the gross margin expectations for the following quarters? Can you kind of talk about why you expect such a significant drop off in gross margin? And maybe talk about how should we think about gross margin moving forward and your seasonality you can speak about?.

Yes. Gross margins are really impacted by the mix of product that we expect to shift in the fourth quarter. The current December quarter had a fairly significant shipment for a high-capacity computing system in Europe, which won't be there in the next quarter. In addition, Nokia revenues are down again next quarter.

Nokia is generally a reasonable level of gross margin, more in line with our average. So we're really impacted by that mix shift from that lost bit of business in Europe in the March quarter.

And going forward, we still see that generally, we should be expecting somewhere in the low 60% range like we've experienced in the most recent quarters, fairly consistently..

Got you.

And with regards to the TAM for the RadHard of $15 million to $20 million, how much would you say you expect to achieve in 2020? And this TAM, do you expect it to grow, expand in the future? Or is this with your long-term TAM?.

Hard to predict what's going to be in 2020. As we've talked about in the past, we had anticipated to already have seen some revenue from the RadHard, which hasn't happened yet. It's just very difficult to predict with launch schedules and funding and what have you.

We certainly anticipate to get something in calendar 2020, we just don't know what it is. As far as the TAM, $15 to $20 million is what we've talked about with the current product value we have. We -- as I mentioned, we've done some testing on the Gemini-1, which looked very promising.

So certainly, we'll be able to add, hopefully, that family to the aerospace portfolio. And if that's the case, that will certainly grow the TAM possibility for that space..

Got you. And just one more question with regards to your path to profitability.

Yes, can you kind of talk about how you're seeing that as we move forward?.

Tim sorry, with the profitability, Ari?.

Correct. Yes. Your path towards profitability..

I don't understand the question..

Yes, yes.

I was just asking about what is your path to profitability?.

Oh, path to profitability. Obviously, RadHard and rad-tolerant are very gross margin products. Generally, we would expect it to be somewhere in the 85% range. And it doesn't take a lot of RadHard/rad-tolerant revenue on a quarterly basis to start breaking even.

In addition to that, once we start shipping the Gemini chip, that also, we expect to be providing a gross margin greater than our corporate average. It's really a matter of being successful in shipping those products and that should lead us to profitability and growth..

Do you have a time line or not at this time?.

We're not giving any forecast on those revenues yet, it's kind of early on the Gemini chip. There is a lot of interest but no one is giving us any kind of a forecast in volume data that we can present to people. In addition, DDA has already talked about the RadHard/rad-tolerant.

It's a matter of getting out there and getting that first order and showing that the part works in space. And again, there's a significant amount of interest on the RadHard and lot of information that we're seeing or interest that we're seeing in the rad-tolerant product..

[Operator Instructions]. We'll take our next question from Kurt Caramanidis with Carl M. Hennig, Inc..

What would be the potential -- if you've got a alpha prospect customer in the defense area.

What would be the potential number of board for a customer like that? Is it a couple? Is it a dozen? Is it 100? Kind of just to get an idea what could an alpha customer have for demand?.

For which product, Kurt?.

For the Gemini.

You're going with the new Gemini, you're looking for design wins, what could some of these larger prospects -- what is their potential? Just in a general sense from single digits to dozens to 100 board, what the kind of opportunity is there in your larger prospects?.

So it really depends on the application in the market first. And what I mean by that is, the volume of Gemini boards for drug discovery versus facial recognition versus e-commerce versus some of these government applications we talk to all differ. So it really depends on the market space and also the customer side.

So for me to generically answer that question, it could be tens that we're shipping to customers to thousands. So it really depends on the space and the customer..

So there is an opportunity for thousands with someone?.

Too early to be talking about quantities. But certainly, if you look at some of the updates that we're looking at, it -- they could be larger. But the answer is, right now, we still need to provide the benchmarking. And we're going to be going through -- actually had our first workshop this week.

We actually went out to a government entity and did our first workshop this -- in fact on Tuesday and Wednesday. And so right now, we still need to prove them that this is a solution beyond just the PowerPoint presentation. And once we get to that next level, and then they'll start talking quantities. But it's too early.

They haven't gotten to that point with us yet..

Okay. Could you kind of know what the customers do? So that kind of a follow up was, the market size for -- across all of these e-commerce and drug discovery and facial recognition, across all of that, I'm assuming -- I shouldn't assume, but the market there, is it tens of million? More than that? Based on what you're saying, I think it's more.

I realize you have to prove it out and that kind of make sure it's working the way you want it to work. With what you've put into it, is it going to be tens of millions or more than that? Or what -- just generally not the -- 10-year down, but just kind of generally speaking..

Yes. So again, whatever numbers we give, it would be our own internal speculation not based off of customer feedback because we're not quite there yet.

I think and within the next couple of quarters, once we have proven our solution, it does what we say it's going to do, and it gives them the benefits and the acceleration and the search that they're looking for, we'll start getting those kind of numbers.

But right now, I don't want to throw numbers out because they certainly have not been provided by the customers themselves..

Okay. Fair enough. And then when you get the benchmarks completed in March, can you pass that along to shareholders? I don't know what this publishing if that's some other technical term, but I think we've been waiting probably a few years.

But when that's done, can we get a release on what the results are?.

Yes. We're looking to do a press release. So it'll be....

So we have to wait months after it's done?.

Hopefully, not. No. I mean, so as we discussed, we've done the testing with the current board we have up to 250 items -- 250-million item database, and we're working on the 1 billion item database now. We're waiting for our new respinned board.

The board, not -- again, if you recall that Gemini is a chip and then it goes on to a board, and the board, the first board has some limitations. So the next board, we'll be receiving within the next week or 2. And then so sometime in February, we'll have it all put together.

So once we have all the testing done, then we anticipate to publish those results fairly quickly..

We'll hear next from Jeff Bernstein with Cowen..

Two questions. I just -- I missed what you said about the qual of the rad-tolerant and RadHard SRAMs, I thought you said Class Q and something else.

I just wanted to know what those were?.

Sure. Yes. So this is our internal qual that we've been working on for the second half of last year. And there's two different -- well, there's a few different levels, but the 2 major levels are what they call QMLQ and QMLV. And we're actually going to be doing -- and those are actually what's called SMD, which is the military drawings.

We're going to be doing under something called SCD, which is source control drawing, which is the QMLQ and QMLV equivalent. So this Q and V are the 2 levels..

Got you.

And those are basically specs in terms of how much ionizing radiation these things can take?.

Exactly. And how much -- what's the TIDs or the total doses, everything? That's correct..

Yes. Yes. Got you. Okay. All right. And then I had a question just on the logistics of supplying product to potential customers and eventually, customers.

What's the minimum lot size that you can order from TSMC?.

For which -- are you talking about....

For the Gemini, yes..

Well, normally, we hold in wafer in the -- the last size is a 25 wafer lot, that's pretty much a minimum. And the maximum, I don't think there's a limitation there..

Got you.

And how long does it take to turn around a 25 wafer order?.

About six months. If we start the wafer from the very beginning, but typically, we have a wafer stage in the various stages of the pipeline. So normally, will be much shorter. But if you start from the first day, day 1, then you will take 6 months..

Okay.

And about how much is the cost of that? So we can assume that you already have one lot size ordered and that you'll have product ready to go?.

We're not allowed to disclose the price of the wafers. But they all vary based on the technology. Obviously, the higher -- the more advanced the technology, the highest price..

Got you.

And so -- but can you say that you have already ordered one lot size, so you have 25 wafers worth of product that you can have for end-use available?.

No. So far, we have our prototype wafer, which does not -- which is another production in lot size limitation. But we do have another wafer, just to get you started, just anticipate some kind of business will come in..

[Operator Instructions]. And we'll take a follow-up question from Kurt Caramanidis with Carl M. Hennig, Inc..

Very nice pronunciation of my name, by the way. How -- just to remind me, how easy or difficult is it to plug the -- I say full, again, I don't know if that's even proper.

But to use this board with existing infrastructure, is that a herculean task? Or is it kind of a plug-and-play or somewhere in between?.

For Gemini board?.

For Gemini board..

Yes. So no, you just -- right. So we did it ourselves here. So for the workshop that we have here in Sunnyvale, we needed, obviously, to have some Gemini boards. And so we have two Gemini boards here at our Sunnyvale facility, and we mounted them in our server here in our server room.

And between just putting them in and just getting them up and running, and it was the first time we did it, took about a day or two with just two guys doing it. So again, this is the first time we've done it. So it's not a herculean effort..

We'll take our next question from George Gaspar with Private Investor..

Could you give me a brief analysis here on your cash position? Could you relate the level of cash that you had at the end of the quarter versus the previous quarter and the beginning of the year -- the end of the last fiscal year? And with that -- then I'll ask you a second part of the question, please..

So at the end of the most recent quarter, we were $73.4 million. At the end of the prior quarter, we were $72.9 million. And at the end of the fiscal year, March, a year ago, we were $70.8 million. So the cash balance has grown each quarter for the last, if I look -- it looks like it's grown every quarter for about 7 or 8 quarters now..

Okay.

And it's kind of interesting that your cash has moved ahead on a quarter-to-quarter basis modestly here, and your loss for the quarter is -- how much you recorded your net loss?.

It's a big number. We were at about $4.6 million..

$4.6 million.

And what was it in the last quarter?.

$1.8 million..

$1.8 million.

So how is it that you can have a $4.6 million loss and increase your total cash -- net cash position? That's got a -- does this lead to the suggestion that you're interworking margin is greater somewhere? How do you get to have more cash losing money the way you are?.

Well, don't forget that there is a significant amount of noncash expenses in your P&L. Your stock-based compensation is running us over $2 million a year. Depreciation and other amortization is probably about $1.5 million a year. So those are all noncash expenses.

Stock-based comp, obviously, we have the required charges based on GAAP, and we're not spending any money there. We typically haven't been buying a lot of equipment. That is the actual use of cash whereas depreciation isn't. We haven't bought a lot of soft for anything for a long time.

The other thing is that, that $2.7 million charge that we incurred this quarter, it's significantly influenced the loss for the quarter, but we're not obligated to pay for that upfront. So there's a fairly significant part of that and accrued liabilities at the end of the quarter.

You'll have to look at the balance sheet, changes in the balance sheet to figure out what's happening with cash..

And then looking ahead in six months out, a year out, do you have any number that you could give us on how much you're going to have to make as an investment into inventory and development to get this board out and into the marketplace?.

Yes. In terms of development, I think we've incurred most of the significant charges at this point. Certainly, there's no significant increase in R&D. Once we start generating sales from these products, we're not going to see SG&A go up much. There'll be a little bit increase in selling expenses for recognitions that we have to pay.

But all in all, the significant charges have already been incurred. The next big thing that we would see is on the current Gemini chip that we're working on right now in R&D, there will be somewhere between $1 million and $2 million charge for mass debts in plus or minus two years..

So when Doug says the current one, he's talking about the next-generation Gemini too..

Yes, okay..

Yes. So I just don't see a big draw on cash at this point. We had a board meeting the other day, we spent quite a bit of time with them going over our strategy with the APU products. And we really don't foresee very significant increases in spending over the near-term for the next year or 2 at this point. We don't see much changing..

And since we have no further questions at this time, this concludes today's question-and-answer session. I would like to turn the call back over to Mr. Shu for any additional or closing remarks..

Thank you all for joining us. We look forward to speaking with you again when we report our fourth quarter fiscal 2020 results. Thank you..

Once again, that does conclude today's conference. Thank you for your participation. You may now disconnect your phone lines..