Ladies and gentlemen, thank you for standing by. Welcome to GSI Technology's second quarter fiscal 2021 results conference call for the quarter ended September 30, 2020. At this time, all participants are in a listen-only mode. Later, we will conduct a question-and-answer session. At that time, we will provide instructions for those interested in entering the queue for the Q&A. 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 have also been asked to advise you that this conference call is being recorded today, which is October 20, 2020, at the request of GSI Technology. Hosting the call today is Lee-Lean Shu, the company's Chairman, President and Chief Executive Officer. And 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 discuss our second 2021 financial results. Throughout the first half of fiscal year 2021, business disruption related to COVID-19 continued to impact our financial results. The second quarter revenue of $6.7 million was at the midpoint of our revenue guidance. Higher SigmaQuad sales in the quarter improved gross margin to 46.7%, exceeding our forecast. Second quarter sales to Nokia were below last year's levels but were up 89% from the prior quarter, offsetting weakness in other consumer segments. Nokia is our largest customer, and earlier this year, they downgraded their outlook for the full year 2020 due to declining demand, as its customers delay their spending plans. We have made significant progress on multiple fronts with the Gemini-I APU. Our revolutionary in-memory processor has continued to amaze us with its adaptability to new applications and our ability to improve the chip's performance through the software. First, our software team has developed Neural Hash, a new algorithm to improve the speed of similarity search for a given accuracy rate. In October, we presented Neural Hash at the 2020 BayLearn conference held by the Bay Area Machine Learning Symposium. Senior representatives from companies including Google, Apple, and Netflix, to name a few, make up the BayLearn organization and the advisory committees providing high level industry exposure for GSI. The BayLearn Symposium's goal is to bring together scientists in machine learning from the San Francisco Bay Area. This event offered exposure to key members of the research community and the technology industry. Second, to simplify writing code for the Gemini-I APU, we developed a code generator and plan to release a compiler stack in early calendar 2021. The compiler will translate code written in Python to APU machine language. This program will allow developers to write code in a high-level language like Python without requiring them to load the chip architecture and the internal logic details. Python is now the most popular programming language with over one-third of the programming market, making this tool key to gaining traction in the market. Third, in the first quarter of calendar 2021, we expect to begin testing the 600 MHz Gemini-I. The goal is to start a qualification and prepare the final product for mass production in the first half of calendar 2021. Lastly, we have made progress on Gemini-II, the second product in our APU family, with a planned launch in late calendar 2022 with orders of magnitude performance-cost improvement over Gemini-I. In closing, I want to highlight the software capability of the APU. In addition to a robust architecture, the APU has a very sophisticated software component. We have a large team dedicated to APU software development. There are four levels to APU software. First, applications. Integration for specific market applications like Biovia for drug discovery, facial recognition, image processing, and HAS search for drug discovery. Second, machine learning algorithm. We developed a dual-level algorithm to enhance APU applications. Third, apps. We developed a library for specific applications as solutions are developed. And lastly, the compiler stack, which I mentioned earlier, is a stack for algorithm conversion, framework support, and simplified low-level code generation. This software capability will give GSI and APU competitive advantage because it will allow us to improve performance without changing the hardware for various applications. With that point, we are beginning to see the rewards of our software investment. The Israeli Ministry of Defense research arm launched the MAFAT challenge, a series of Data Science prize competitions. The most recent contest, the MAFAT Radar Challenge, a machine learning competition to distinguish between humans and animals in radar data. The event drew over 1,000 participants and more than 4,300 submissions. I am pleased to announce that MAFAT notified us this week that the GSI algorithm is at the top of the leaderboard. In the next few weeks, MAFAT will validate the leaders' eligibility and officially announce the winners. If GSI is pronounced the official winner, this prestigious award could bring attention and heighten the credibility of our software achievements, particularly with the Israeli projects that we are currently working on. The entire GSI team is working hard to deliver on our goal of monetizing the investment we have made in the APU. Our timeline has been extended due to the COVID-19 pandemic, but we are making material progress on the technology and are marketing to build awareness. I am grateful for everything that my team has accomplished, and I thank you for your support, first of all, GSI shareholders. Now I will hand the call over to Didier, who will discuss our business performance in further detail. Please go ahead, Didier.

Thank you, Lee-Lean. Starting with the sales breakout for the second quarter of fiscal 2021. Sales to Nokia were $3.4 million or 51.7% of net revenues compared to $5.3 million or 45.2% of net revenues in the same period a year ago and $1.8 million or 26.9% of net revenues in the prior quarter. Military defense sales were 26.9% of second quarter shipments compared to 23.4% of shipments in the comparable period a year ago and 30.1% of shipments in the prior quarter. SigmaQuad sales were 65.4% of second quarter shipments compared to 63.5% in the second quarter of fiscal 2020 and 46.3% for the previous quarter. For the remainder of fiscal year 2021, we continue to anticipate a challenging business environment related to COVID-19 restrictions. Changes in customer buying patterns, communication constraints with our customers, the postponement of investment, and the restricted activity of our sales force and distributors may impact demand and our ability to close sales across customer segments. With this backdrop, our new product sales process is taking longer than usual. Selling new products requires a longer cycle than selling established products. Ideally, we conduct sales meetings in person and spend the time educating customers on how our new product will change their current practices. Without face-to-face meetings, our training workshops, the pace moving the sales process forward has slowed. The targets for radiation-hardened SRAMs are mostly national assets and top-secret applications. Sales communications in this channel slowed due to the lack of access to secure communications facilities. In late calendar 2020 or in early 2021, once funding is released, we anticipate radiation-tolerant SRAM orders for imaging satellites and space applications. Similarly, we are making progress on Gemini-I customers on several fronts as prospects test the solution. We are working closely with a few potential large customers now. That said, given the business challenges to educate and train new customers, we now anticipate design wins and initial sales of Gemini Leda boards into calendar 2021. On the marketing front, GSI has taken steps to raise awareness in the relevant sectors for the Gemini-I APU and our company. We are engaging with industry analysts to gain exposure to leading providers of market intelligence for the technology and telecommunication markets. Our team is expanding its participation in high profile industry events and conferences where GSI can gain recognition as a leading developer of a revolutionary solution for AI computing and have exposure to leading companies in this sector. Globally, we are redesigning our market-facing collateral, including our website, to align our messaging and image with the identity of technology leaders. All of these efforts are to support our sales teams as they build new relationships and bridge the GSI reputation from being a leader in the SRAM memory market to an innovator in the AI market. I'd like to hand the call over to Doug. Doug, go ahead, please.

Thank you Didier. We reported a net loss of $5.2 million or $0.22 per diluted share on net revenues of $6.7 million for the second quarter of fiscal 2021, compared to a net loss of $1.8 million or $0.08 per diluted share on net revenues of $11.7 million for the second quarter of fiscal 2020 and a net loss of $6.1 million or $0.26 per diluted share on net revenues of $6.6 million for the first quarter of fiscal 2021. Gross margin was 46.7% compared to 55.9% in the prior year period and 46.1% in the preceding first quarter. The changes in gross margin were primarily due to changes in product mix sold in the three periods. Total operating expenses in the second quarter of fiscal 2021 were $8.3 million, compared to $8.5 million in the second quarter of fiscal 2020 and $8.7 million in the prior quarter. Research and development expenses were $5.7 million compared to $5.8 million in the prior year period and $5.8 million in the prior quarter. Selling, general, and administrative expenses were $2.6 million in the quarter ended September 30, 2020, compared to $2.7 million in the prior year quarter and down from $2.9 million in the previous quarter. The second quarter fiscal 2021 operating loss was $5.2 million, compared to $1.9 million in the prior year period and $5.7 million in the prior quarter. The second quarter fiscal 2021 net loss included interest income and other expense net of $16,000 and a tax provision of $62,000, compared to $210,000 in interest and other income taxes and the tax provision of $55,000 for the same period a year ago. In the preceding first quarter, the net loss included interest and other income of $106,000 and a tax provision of $487,000, primarily resulting from the settlement of a tax audit in Israel for fiscal years 2017 through 2019. Total second quarter pretax stock-based compensation expense was $653,000, compared to $642,000 in the comparable quarter a year ago and $755,000 in the prior quarter. At September 30, 2020, the company had $56.1 million in cash, cash equivalents and short-term investments and $8.7 million in long-term investments, compared to $66.6 million in cash, cash equivalents and short-term investments and $4.1 million in long-term investments at March 31, 2020. Working capital was $59.2 million as of September 2020, versus $70.9 million at March 31, 2020 with no debt. Stockholders' equity as of September 30, 2020 was $82.2 million, compared to $89.6 million as of the fiscal year ended March 31, 2020. For the upcoming third quarter fiscal year 2021, our current expectations are net revenues in the range of $6 million to $7.2 million with gross margin of approximately 41% to 43%. Operator, at this point we will open the call to Q&A.

We will take our first question from Denis Pyatchanin on behalf of Raji Gill. Please go ahead.

Hi guys. Thank you for taking my call. So I want to ask a couple of questions. For the first one, about the Gemini-I APU. Could you talk a little bit about just how feedback from some early adopters? And then maybe when we should expect to see revenue from this product and the revenue opportunity? I mean how is it different from other AI startups doing in-memory processing as well?

So the feedback has been really good so far. As I mentioned, we have a couple larger companies that we are starting to talk to now that are new in the past calendar quarter. And then we have been working with some of the government entities, along with some drug discovery and facial recognition companies. All the feedback has been very good. Now as far as how do we compare with the competitors? What we have seen so far is that everyone is concentrating their efforts on the training. That's now where we are. We are concentrating our efforts on similarity search. So for us, we don't care who does the training, whether it's Nvidia or a startup or Intel. For us, all we need is a vectorized database and that's where we take over. So we are really not competing with those folks. If you look at how those search functions are done today, it's mostly done in CPUs. So we are really replacing some old servers using standard CPUs in the search function.

Got it. Thanks. And then is there anything you can give us about the Nokia business going forward? Kind of the impact on the gross margins in terms of the product mix? Or any other color you can provide on the Nokia situation?

So Nokia is tracking about along the lines that we have been predicting for a few quarters now. We have predicted that the March quarter and the June quarter would be down and that they would be recovering in the second half of calendar 2020. And that's what happened. So the two quarters were down, and then we recovered this quarter. The forecast we have seen so far continue to bring us back to the levels we are at now, so more of the run rate kind of levels.

Great. And then just really briefly, is there anything that you guys have been thinking about doing about the whole kind of the RadHard customer issues with the marketing that you can't see them in person because of COVID? Is there kind of any plans about trying to do something virtually or differently about that?

So virtually, it doesn't work. It has to be some kind of secure communication. And so Zoom is out. Phone calls are out. Emails are out. There are really two ways. It's face-to-face in a secure facility or there are some encrypted secure communications that are also within buildings. They are not anything that we have. So we had a few face-to-face meetings. One of our team members had to travel to New Mexico to do that. But certainly, it's hampering that effort on the RadHard. As we discussed in the script, the radiation-tolerant generally does not have a lot of those top-secret notations on them. So we are starting to see some movement in the RadHard and anticipate seeing our first order, if not this quarter, certainly it should hopefully happen next quarter assuming the funding has been released.

All right. Thank you very much. That was all for me.

Thanks, Denis.

We will now take our next question from Jeff Bernstein from Cowen. Please go ahead.

Yes. Hi guys. I just want to make sure I heard some of the timing right here. So the 600 megahertz board for the APU I is in testing now? And you see sales in H1 of 2021? Is that right?

So the solution is two-pronged. So it's the chip and the board. Right now we have the 400 megahertz chip with what we call the Leda-G board. That's what we have today and that's what customers who have tested our solution are looking at in all of our demos. The next rev chip, which would be the 600 megahertz chips, we will see it at the beginning of 2021. In fact, we should have it in hand by January. Now that solution also requires a little bit different Leda board, especially called Leda-E board. The Leda-E board, we should have some of those in hand by next month in November. So we won't start any testing until January at the soonest.

Okay. And that's sort of an interim speed bump. You don't actually really have to have that to make some sales before that?

Correct. So if you look at some of the solutions that we are going after, the facial recognition, the drug discovery, and some of these government and military applications we are looking at, the 400 megahertz solution is fine for them. Certainly, we like to see a bump in performance even further than what we have. And that's why we are offering the 600 megahertz. But like you said, there are applications where the 400 megahertz is fine.

Yes. Got you. Okay. And then on the Gemini-II, I think, did you say May 2022, you will have that in hand? Or what was that scheduling?

Yes. We said at the end of 2022, calendar 2022, we will have a chip on hand.

So the end of calendar 2022?

Yes.

Got you. Okay. Great. And then in terms of the design cycles for RadHard and some of the just military applications for APU, what does that look like? So if you are talking to a guy like Mercury Computer that specializes in using off-the-shelf signal processing for CQ guide kind of applications and they go, wow, this is great, super fast, right, recognition of targets and things like that, radar signals, etc. We want to use that. How long is it before they can actually be buying that and have it in a subsystem that is drone or UAV or whatever it is?

So that's a good question. It's hard to predict. So the first thing we need to do is, obviously get the solution in their hands. So we have two folks who have already purchased in kind of that market segment you just mentioned, the military area have bought two Leda boards already, and we anticipate, we don't have the orders yet, but we anticipate that we should sell five more Leda boards to that segment this quarter. These are all different customers within that realm. They need to get the board and start playing with it. Now, we are showing them how to write their own microcode, and as Lee-Lean mentioned earlier in the talk, we have a code generator already. But a full-blown compiler stack isn't going to be ready until January. So if they are looking to write their own, they can get started now. If they are looking to be able to take one of their algorithms or software and feel kind of throw it into our compiler at a high-level language, that's going to be early next year before they can do that. So it's a little hard to predict the timing at this point.

Great. Thank you.

At this time, there are no further questions, so I would like to return the conference to you for any additional or closing remarks.

Thank you all for joining us today. We look forward to speaking with you again when we report our third quarter fiscal 2021 results. Bye.

That concludes today's call. We thank you for your participation. You may now disconnect.