Lee-Lean Shu – Chairman, President and Chief Executive Officer Douglas Schirle – Chief Financial Officer Didier Lasserre – Vice President-Sales.

Kurt Caramanidis – Carl M. Hennig, Inc. Jeff Bernstein – Cowen.

Ladies and gentlemen, thank you for standing by. Welcome to GSI Technology’s Second Quarter Fiscal 2018 Results Conference Call. 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, October 26, 2017, at the request of GSI Technology.

Hosting the call today is a Lee-Lean Shu, the company’s Chairman, President and Chief Executive Officer. With him are Douglas Schirle, Chief Financial Officer; and Didier Lasserre, the 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. Today, we reported second quarter net revenue of $9.6 million and gross margin of 50.4%. Both were below the range of guidance that we provided earlier in the quarter. During the fiscal second quarter, we continued to see slowness in our primary telecommunication and the networking markets.

That said, we remain very excited about all product development and marketing efforts for our newest market segment, our high-performance SigmaQuad radiation-hardened SRAM products, targeted at aerospace and defense applications, and our patented in-place associative computing technology, or APU.

We now have packaged parts of our SigmaQuad radiation-hardened SRAM products. Once we finish processing these parts, they will be available to ship as revenue-generating samples during the third quarter of fiscal 2018.

Starting this week, we are beginning the qualification process for our MIL-PRS-38535 Class V equivalent product for shipments in calendar 2018.

We are on schedule to complete the design of our initial APU product by the end of January 2018, and we continue to be focused on expanding our overall market position in high-speed SRAM and the low-latency DRAM segments, complemented by our RadHard SRAM products.

We continue to receive inquiries about our patent in-place associative computing technology and the intellectual properties that is under development, which focus on large emerging markets such as big data applications, computer vision and cybersecurity.

At the recent In-Memory Computing summit, we presented how the APU is well-suited to Natural Language Processing.

We were also recently invited to make a technical presentation on the APU at one the leading aerospace research and the development centers and were also invited to present APU architecture at Stanford University as part of their EE Computer Systems Colloquium.

As a reminder, we obtained our APU technology through our acquisition of MikaMonu in November 2015. This technology is for applications in evolving new markets such as big data, including machine learning and deep convolutional neural network, computer region and cybersecurity. We currently hold 15 U.S.

patents and a number of pending patents for our in-place associative computing technology. The technology changed the concept of computing from serial data processes, where data is moved back and forth from the processor to the memory to parallel data processing, computation and the search directly in the main processing array.

Whilst still in development, we are currently marketing new product sets data from this technology to a variety of users. And we believe we will have an active base of potential customers once we bring this new product to market.

This technology will readily improve the computation and the response times that are currently and will continue to be required in a variety of big data applications. Overall, we will remain on time to execute to our primary business objective to bring our RadHard products to market and completing our APU design in the very near future.

We anticipate both products to bring us into growing market segments and broaden our customer base in calendar 2018..

We reported a net loss of $1.7 million or $0.08 per diluted share on net revenues of $9.6 million to the second quarter of fiscal 2018 compared to net income of $626,000 or $0.03 per diluted share on net revenues of $13.4 million for the second quarter of fiscal 2017 and a net loss of $1.5 million or $0.07 per diluted share on net revenues of $10.7 million for the first quarter of fiscal 2018, ended June 30, 2017.

Gross margin was 50.4% compared to 55% in the prior year period and 52.4% in the first quarter. Second quarter fiscal 2018 operating loss was $1.8 million compared to an operating loss of $1.5 million in the prior quarter and operating income of $380,000 a year ago.

Total operating expenses in the second quarter of fiscal 2018 were $6.7 million compared to $7 million in the second quarter of fiscal 2017 and $7.1 million in the preceding first quarter. Research and development expenses were $4.2 million compared to $4.3 million in the prior year period and $4.3 million in the preceding quarter.

Selling, general and administrative expenses were comparable year-over-year at $2.5 million compared to $2.7 million in the quarter ended September 30, 2016, and down sequentially from $2.8 million in the preceding quarter.

Total second quarter pretax stock-based compensation expense was $508,000 compared to $478,000 in the prior quarter and $487,000 in the comparable quarter a year ago.

Second quarter fiscal 2018 net income included interest and other income of $103,000 and a tax provision of $49,000 compared to $92,000 in interest and other income and a tax benefit of $154,000 a year ago. In the preceding quarter, net loss included interest and other income of $98,000 and a tax provision of $81,000.

In the second quarter of fiscal 2018, sales to Nokia were $2.8 million or 29% of net revenues compared to $4.5 million or 42.3% of net revenues in the prior quarter and $5.7 million or 42.8% of net revenues in the same period a year ago.

Direct and indirect sales to Cisco Systems were $506,000 or 5.2% of net revenues compared to $852,000 or 8% of net revenues in the prior year – in the prior quarter and $936,000 or 7% of net revenues in the same period a year ago.

Military/defense sales were 24.4% of shipments compared to 24.5% of shipments in the prior quarter and 16.1% of shipments in the comparable period a year ago. SigmaQuad sales were 39.4% of shipments compared to 51.1% in the prior quarter and 57.2% in the second quarter of fiscal 2017.

At September 30, 2017, we had $49.6 million in cash, cash equivalents and short-term investments; $14.6 million in long-term investments; $56.3 million in working capital, no debt; and stockholders’ equity of $85.6 million.

Looking forward to the third quarter of fiscal 2018, we currently expect net revenues to be in the range of $10 million to $11 million. We expect gross margin of approximately 53% to 55% in the third quarter. Operator, at this point, we’ll open the call to Q&A..

Thank you. [Operator Instructions] We’ll take our first question from Kurt Caramanidis from Carl M. Hennig, Inc..

Hi guys..

Hi, Kurt..

First question on the RadHard.

Can you maybe just talk about the level of interest if that’s changed since the last quarter? And are we looking at any revenues – I’m going to use calendar quarter, I think most people probably understand that – any revenues in calendar Q4 or is it mainly starting in calendar Q1 of 2018? And what are we may be looking at roughly dollar-wise in calendar Q1 if that one is to start?.

Kirk, it’s Didier. So the – yes, there is a level of interest, a high level of interest in both the original offering that we have, which is the 288-megabit SigmaQuad family.

And also, I think I mentioned in the last call that we were also accelerating the introduction of our 144-megabit SyncBurst and NBT device as well for a little bit different application within that sector. So the 288-megabit is the family that Lee-Lean mentioned that we had packages in-house.

And then starting this week, we will be starting the qualification for the full production Class V equivalent. On that family, we are expecting a handful of parts, engineering samples being sold in this quarter. So the answer is yes. So the level of interest is still there.

There has been some freezing in some of the funds from IRAD, which has pushed out some of the anticipated revenues from that family. So we were expecting a little bit more this quarter, but because of those funds being frozen temporarily, it has pushed out some of the revenue.

So we will have somewhere, if I were to guess, about $100,000 in revenue this quarter and then ramping starting the first calendar quarter of next year..

What are you thinking in that quarter?.

It’s hard to tell. We’re still dependent on funding. It’s a little bit different model than we’re used to with our classic telecom and classic military customers.

Now this requires a lot of funding from the government, and it’s not clear that the funding – how much is going to be released for the calendar first quarter or how much is going to actually fall into the following quarter. So it’s hard for me to predict at this point..

Okay. That’s fair. And then just overall, on the stages of the APU unit, it’s taken about roughly two years to design it and then to review.

Is that – what’s the time period when – in January, get the design done, then it goes to TSMC? Can you talk about the time line it takes to manufacture the chip? And then from there, kind of what you do if you test it and that kind of time frame now that we’re kind of getting to those times?.

Yes, once we release the design to TSMC, it normally takes around six to eight weeks to generate the silicon. So once we generate the silicon, it will probably take another few weeks, two, three weeks to do the packaging. After that we can start testing. This product will be extending the fueling out to the car instead of the package parts.

So hopefully, we will do another testing to put the parts into the car. And hopefully, if everything work out well, we should be able to send now the car hopefully by the mid of third quarter, next calendar year..

Actually, Kurt, we’re still expecting to get those initial package parts back, sometime end of July or probably around the middle of April at the latest. So there’s been no change currently in our expectations than what we previously talked about regarding that product..

End of March, Doug..

End of March..

Yes. And….

March and middle of like the April time frame..

So how about the difficulty – as we – is the design by far more difficult than the manufacturing and the testing? Or how do you think about – have we climbed a mountain when we got to the end of the design and then its chances of success are better when that’s done? Or how do you think about the three phases and difficulty?.

Of course, the design, we don’t have any learning curve before. This is the very first chip we start to design. So hopefully, we’ll have some sort of learning curve. And once the silicon come out, that’s a debugging process. Again, that’s also the very first time that we are targeting the debugging testing process for the product.

So you’ll come down to how well the design is. And hopefully, we have a very thorough design, and there’s going to be less debugging process. But anyway, overall, I will think sometime second half of 2018, we should have the product out for customers to do the testing for the ongoing development..

Okay. Great. And then when you’ve got – these presentations you’ve got available online, it sounds like you’ve got an aerospace space one coming and then another one.

Or if they’re not live, can you get them on the website recorded to some extent?.

Well, I know that the presentations at standard – at Stanford will be recorded. And I believe that they’re going to give us a copy of that. So yes, we can try to get that available on the website somehow..

Actually, if you go to Stanford University EE380, that’s the EE Computer Systems Colloquium, if you search for Stanford University EE380, you will see the agenda and you can get the webcast from there..

Okay. It’ll be good to have it on the website too, yes..

But the others were specific meetings with the customers. The aerospace one that Didier talked about, that’s just a meeting. It’s not necessarily a presentation. But there was a presentation that was made at In-Memory Computing seminar that we were presenting at this week.

I don’t know if that was recorded or not or if there’s anything available regarding that..

Okay, no problem. Finally – I apologize, Dr. Akerib talked about in January that there wasn’t really going to be any major developments this year in kind of the APU area, which I have not seen, and he seems to be on track on that.

What is he thinking about 2018? Does he still view your position as a top position and a pretty major advancement compared to current products?.

Yes. I mean, we still believe that we’re significantly better from search perspective and aspects of search. And then Lee-Lean and Didier are more technical than I am. But when we look at this product that – and the parts it’s competing against, this is significantly more encompassing in terms of what it can do.

So we made this one presentation on Natural Language Processing. The more and more we get into that and the more and more customers we talked to, we see there is a much greater need in a lot of areas than we originally anticipated for this part.

So if anything, I think we would say that we’re seeing more opportunities than we would have initially expected two years ago when we got into this..

Right. And Kurt, let me add to that. So we’ve been flying kind of underneath the radar until we actually see silicon. And then we’ll certainly be more aggressive in implementing the device.

But the promotion we’re doing now as we talked about at some of these shows – and just the show that we had this week, as Doug mentioned, the In-Memory Computing, we had a few companies that are large companies, approach us after we presented and wanted to have much further in-depth conversations with us.

Because – in fact, one of the companies, the gentlemen there were highly technical and they were slated from their company to be looking for technologies along what we’re doing and they had not found anything that remotely touch what we were doing. So they were very excited.

So certainly, again, we haven’t – we’ve been a bit under the radar, but these shows have gotten us out into the technical community, and we have a lot of interest out there..

Great, sounds good luck for the next quarter..

Thanks Kurt..

[Operator Instructions] And next we’ll hear from Jeff Bernstein with Cowen..

Just wanted to follow up along those lines on the APU.

So have you guys done emulation already on the part?.

We – no, we don’t hardware emulation but we do have a software model for the APU. So actually, we have studied some academics in the university. Actually, the professors were paying such thinking at getting our software model so that they can start simulating. So they do have interest in them..

So you don’t have any kind of new metrics in terms of – you’ve talked in the past about what you thought the improvement in processing speed was.

So that hasn’t really changed?.

Correct..

Got you. And then talk a little bit about the software side of this. I guess, I’m assuming that people are going to have to program things differently in order to take advantage of the parallelism. Are you going to supply that software.

And I guess, if there’s – are there any new software tools that need to be written for this? And how long will that take? And how do you think that will play out?.

No, we – actually, we have a pretty big team of software group. Basically, what we will do is we will do a library, okay? So people can still use a top-level language to call the functions of what the chip’s supposed to do. Basically, it’s concurrent with them. They can save, modify, addition, just like how they use it in the typical programming.

And underneath, it will be at the function we built, so they can use that at the library or use their common function algorithm – for our search and keying in that common use of function at the library to go with the chip..

Great.

So when you ship the cards, then people should be able to start using them right away?.

Correct..

And Jeff, let me add to that. So we’re – as Lee-Lean said, we have a large team that’s going to be writing the libraries for us. Obviously, we’re going to write all the major tasks and functions, but there will always be other libraries that are needed.

Doug – or I’m sorry, Lee-Lean mentioned earlier that recently we had met with a leading aerospace company about the APU. And they certainly wanted to learn about the APU, the physical, the hardware portion of it. But the other part of the meeting was talking about the software.

So we’re going to be getting back with them to start training their folks to be able to program their own library or write, I should say, their own libraries. And we’re anticipating that a lot of folks, especially in the government and military sectors, will be writing the libraries for our parts.

And what’s nice about that is that the majority of them will be available for all users. And what I mean by that is, obviously, some aspects of the government, whatever they write, they want to keep in-house and not share. But the majority of the libraries that will be written by these entities will be available for use by all users.

So the nice thing about getting this board out of the card that Lee-Lean said, is that once we get the cards out, we’re going to put them in not only academia’s hands, but also some of these other folks that will start writing the libraries for us that will help expand the software portion..

Got you. Okay. And then I think I missed when you were going through revenue concentrations, Nokia.

And can you just talk about anything that’s going on there in terms of some of their new products and where you guys think you stand?.

Sure. So we’ve talked about the last couple of quarters that they’ve been having some inventory correction. They took a hard correction this past quarter on the older systems, the ones that use our 72-megabit SQ-III. Certainly, the forecast they give us with this quarter shows them coming back up.

Now as far as the new systems, as I’ve mentioned in the past, I’m not – I don’t think it was the last call, it was a couple of calls ago, we talked about a mid-life kicker platform based off of the same core router and edge router that we’re shipping our 72-meg in. So that mid-life kicker will be used at a 144-megabit SQ-III.

So this past quarter, we started shipping some of those. It wasn’t super high-volume yet, but we’re seeing that forecast slowly coming up. And so that’s really the kind of the next generation for us is that mid-life kicker on the Nokia 7750 and 7950 platforms..

Great. Okay.

So you are in there and will have some business from there going forward?.

On the mid-life kickers, that’s correct, yes. And it’s already started. It’s mostly prototype-ish quantities, hundreds of pieces a quarter, but that hopefully will ramp..

Okay, perfect. Thank you..

Thanks Jeff..

[Operator Instructions] I show no further questions in the queue, gentlemen..

Thank you all for joining us. We look forward to speaking with you again when we report our third quarter fiscal 2018 results. Also for those attending, we will be presenting at the upcoming 10th Annual IT – LD Micro Main Event Investment Conference, Los Angeles, on December 6. This presentation will be webcast as well. Thank you..

And that concludes today’s conference call. We thank you for joining..