GSI Technology Inc (GSIT) 2019 Q3 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to GSI Technology's Third Quarter Fiscal 2019 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 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, January 31, 2019, 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 today's call to review our third fiscal quarter 2019 result.

I am very pleased with the performance of our core business, which exceeds our forecast primarily due to better-than-anticipated SRAM sales to our largest customer, which increased 11% sequentially and 72% year-over-year.

Net revenue for the quarter was $14.7 million, above the high end of the guidance that we provided from our second quarter call of $12.8 million to $13.8 million.

With gross margins of 68%, [rose] above our guidance of 62% to 64%.

Doug Schirle, our CFO, will provide additional insight into the factors that drove our strong revenue growth and the margin improvement in his commentary.

Earlier this month, we provided an update on the APU, our in-place associated processing solution for artificial intelligence applications.

We are currently testing packaged units of the first silicon wafers.

These initial tests have identified some bugs, which have been fixed now.

In the meantime, our hardware and software teams are working on demo board with the APU and determine if their performance are ready for testing by our selected alpha customers.

At this time, we [intend] to receive the [next] silicon pack in a few months, and we anticipate being able to perform [payments] for beta customer later in the year.

In early November, a member of our APU team hosted a panel on video research -- on video search at ODSC West conference.

ODSC is a global data science community dedicated to helping foster the change of innovative ideas in [urban] data science.

The panel, led by our Director of Data Science, discussed that video search is the next frontier of search.

In the panel discussion were experts in the field from eBay, Google, Walmart Labs, Wayfair and the [Carife].

Largely driven by video content, video search is ideal for online retail.

This search mechanism allows you to [request] text entered on a keyboard with image captured on your phone to search for objects such as clothing or furniture.

Similarity search use a technique called distance metric learning, in which learning algorithm measure how similar or related objects are.

GSI's APU is ideal for similarity search because its unique design enables very fast computation speed with improved accuracy and scalability for very large data set, which is key to improve similarity search capabilities.

We are very excited about the similarity search applications for our chip.

AI adoption [individual] is forecast to grow very rapidly in the next 5 years, with video search being one of the fastest-growing segments of similarity search.

In software simulations, the APU has demonstrated the ability to increase the rate of computation by orders of magnitude with great accuracy while reducing power consumption significantly.

This kind of performance can potentially transform the online retailers' capabilities and greatly improve the online customers' experience.

Other applications of similarity search are in drug discovery, drug screening and genomics.

In drug discovery, the objective is to develop safe and effective drugs, cheaper and faster.

One area of exploration has focused on High Performance Computing.

The goal is to maximize the use of computing through the entire drug discovery process, from the early stage of identifying and designing [seek] molecules to target disease all the way to predicting outcomes of a clinical trial.

Simulation with the GSI's APU for this application have shown to improve search speed by orders of magnitude, with the reductions in query time from as much as 10 minutes to as little as 300 milliseconds.

The APU's highest speeds and increased accuracy in similarity search can potentially lower drug discovery cost.

[Any] important goal for research organizations depend on the funding.

Drug discovery is very costly with a high failure rate.

According to a study from the Tufts Center for the Study of Drug Development, the average cost to develop a new drug is about $2.6 billion.

Only 10% of new drugs win approval, and those few drugs that do get approved can take at least 10 years to come to market.

Technology that can reduce cost, increase drug efficacy and safety and increase speed to market can potentially save billions of dollar.

As a result, the APU is getting attention of leading pharma and genomics players.

I am very pleased with the work our team has done to advance the APU and to increase our visibility with potential key customers.

It is still early and we have a lot of work ahead of us, but I remain confident that the APU will be a successful product for GSI.

As I have stated before, to offset our mature legacy networking and telecommunication market, the company's resource are focused on the development and the launch of new products.

We intend to continue penetrating markets that require high-performance memory technologies for our core products like our SigmaQuad (inaudible), which are recognized as leaders in the industry with their combination of capacity and the performance and the (inaudible) are unequaled by any competitor.

We are waiting on final testing of our SigmaQuad radiation hub in the SRAM product for our first customer.

We packed the first shipment of our prototype SigmaQuad radiation-hardened SRAM in the first quarter of fiscal 2019, which with production shipment of fully qualified product later in 2019.

Our financial performance this quarter shows that when our higher-margin sales increase, we can deliver significant leverage in earning per share growth and improve cash flow from operation.

As we scale with new products, this operating leverage will continue to improve and our cash reserve should grow, providing the company with resources to advance opportunities for existing product line and develop new products.

I am proud of the work my team has accomplished and remain confident that GSI is well positioned to return to growth and improve shareholder value in the long run.

I will now hand the call over to Doug for [this year's] overview of our financial results.

Doug?

Thank you, Lee-Lean.

We reported 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 compared to a net loss of $1.5 million or $0.07 per diluted share on net revenues of $11.1 million for the third quarter of fiscal 2018, and a net loss of $351,000 or $0.02 per diluted share on net revenues of $12.8 million for the second quarter of fiscal 2019 ended September 30, 2018.

Gross margin was 68.3% compared to 51% in the prior year period and 62.6% in the preceding second quarter.

Strong revenue growth in the third quarter was driven by increased sales to our largest customer, the balance of the shipment of a onetime order for a university supercomputer build-out in Europe and an increase in sales in our military segment.

These factors drove an increase in gross margins, which benefited from a favorable product mix.

Total operating expenses in the third quarter fiscal 2019 were $7.8 million compared to $6.7 million in the third quarter fiscal 2018 and $8.4 million in the preceding second quarter.

Research and development expenses were $5.2 million compared to $4.2 million in the prior year period and $5.8 million in the preceding quarter.

Research and development expenses in the prior quarter included an expense of approximately $1 million related to a nonproduction mask set for our initial APU product.

Selling, general and administrative expenses were $2.6 million compared to $2.5 million in the quarter ended December 30, 2017 and $2.7 million in the preceding second quarter.

Third quarter fiscal 2019 operating income was $2.2 million compared to an operating loss of $1 million in the comparable period a year ago and $394,000 in the prior quarter.

The third quarter fiscal 2019 net income included interest and other income of $96,000 and a tax provision of $70,000 compared to the same period a year ago, in which net loss included $99,000 in interest and other income and a tax provision of $590,000.

In the preceding quarter, net loss included interest and other income of $145,000 and a tax provision of $102,000.

Total third quarter pretax stock-based compensation expense was $592,000 compared to $535,000 in the comparable quarter a year ago and $552,000 in the prior quarter.

In the third quarter of fiscal 2019, sales to Nokia were $6.2 million (sic) [$6.6 million] or 45.2% of net revenue compared to $6 million or 46.6% of net revenues in the prior quarter and $3.9 million or 34.7% of net revenues in the same period a year ago.

Military/defense sales were 16.7% of shipments compared to 16.4% of shipments in the prior quarter and 20.7% of shipments in the comparable period a year ago.

SigmaQuad sales were 69.7% of shipments compared to 66.4% in the prior quarter and 54.5% in the third quarter fiscal 2018.

At December 31, 2018, the company had $57.1 million in cash, cash equivalents and short-term investments, $10.2 million in long-term investments, $65.5 million in working capital, no debt and stockholders' equity of $91.2 million.

Looking forward to the fourth quarter of fiscal 2019, we currently expect net revenues to be in the range of $11.6 million to $12.6 million.

We expect gross margin of approximately 60% to 62% in the fourth quarter.

Operator, we'll now open the call to Q&A.

(Operator Instructions) And we'll take our first question from Brian Swift with Sutter Securities.

Can you clarify one thing for me?

In the press release, it said your sales to Nokia were $6.6 million, and I think I heard you say that they were $6.2 million.

I just want to know which one was...

No, in the current quarter, it was $6.6 million.

Yes, you said $6.2 million.

But it's $6.6 million.

Brian, it's $6.6 million.

(Operator Instructions) And we'll take our next question from Kurt Caramanidis with Carl M. Hennig, Inc.

It's probably the best quarter you've had in about 7 years if I can remember, looking at it correctly, so nice quarter.

On the APU, do you not really test anything until it's debugged, until the second -- I think you referred to second silicon in that presentation, till that's back?

I'm sorry, can you repeat that?

Do you test it?

I'm wondering if you have any idea if the -- if it's looking like the simulators?

Or do you not know until you have debugged in -- around the second silicon as far as how it's performing compared to what you were thinking prior?

Right.

So we've had the assembled units in hand since November.

And for the first 2 months, we have been looking at the part on a tester, like on a memory tester, so on a component level.

And so that's where we've noticed some of the bugs that we have.

So very recently, within the last week or so, we have now mounted the APU onto a board.

And this is a board that would be one that we would use for demos, so it would include an FPGA and other components as well.

So for the first 2 months, we -- the testing was done on a component level.

And now we are actually running software and algorithms through the device as much as we can with the limitations of the bug.

So we are testing it on both a component level and also on a board level.

Okay.

So what kind of -- in the next quarter, you might have an idea how it's looking on the board?

Or what -- how -- what kind of time line or...

So this is first silicon.

And with this first silicon, we're still hoping to demo one of our alpha guys that we've mentioned in past press releases.

But for the majority of the folks, we're really waiting for second silicon.

Second silicon, we'll be fixing the bugs that we've identified.

That should be done within the next couple weeks.

Of course, at that point, you have to go through a series of changing the mask sets and [starting] new wafers and assembling up the parts.

So it will be sometime, I would say, end of May at the soonest, maybe beginning of June before we have devices in hand, at which point, we're going to look at them in 2 different places in parallel.

One, we're going to put them back on the testers, and then we're also going to mount those on demo boards immediately.

What we're hoping is that we'll be able to start doing a more broad-scale demo-ing sometime in the calendar third quarter of this year.

Okay.

Great.

Do you have to do another $1 million mask?

Or is this using the same masks that you had, you're just tweaking it, [if that's proper]?

Just certain levels will be tweaked.

It won't be anywhere near $1 million.

And just so you recall, that initial mask set we charged to R&D expenses.

Future masks on this will be charged to cost of goods sold over a 12-month period.

So you'll never see any big impact like $1 million on this, or you shouldn't.

Okay.

And then Lee-Lean mentioned some drug -- other drug companies.

So you have the one deal, you have other large drug companies that have expressed interest?

Or what -- or how -- what is -- where is that at?

So certainly, we have -- as we've spoken about in the past, we have the lead alpha customer.

They want early access, so they're willing to look at material that they know have some bugs in them.

We're limiting how many people we engage with on first silicon, obviously, because it's very resource intensive to do that because there's workarounds and what have you.

But there have -- we certainly have seen interest from other folks in the same market, but we're going to put them in the second silicon group.

Great.

And then the rad-hard.

So you're anticipating a prototype this current quarter.

If that testing goes well with your formula, is that then -- would that determine the timing of a sale, a real production sale?

So we -- right.

So we actually received an order last quarter, in the December quarter, for the first prototype.

And then this week, we actually received a second order also for a prototype quantity for a second program.

These are ES units because we have not finished the qualification internally.

We do feel the recipe we have now does meet the requirements.

We still have a few tests to conduct before we can for sure know that.

But the prototypes are going to be sent for these 2 programs for their initial qualification.

And so what we're looking at now is probably finishing up the qualification sometime this quarter.

It might spill into early next quarter.

And then during that time, we're waiting for the customers to finish their qualification with the prototype units that they're ordering now.

We'll have those shipped out within a month, maybe 1.5 months.

Okay.

Then the production orders would follow once everybody is all through qualification?

Correct.

Okay.

Sorry, last one.

Any outreach in the next quarter or so, whether it's an AI conference or any other investor conferences?

Just curious.

Yes, we'll be going to the ROTH Conference and then our IR firm is looking at non-deal roadshow-type activities.

So we'll be out a couple times in the next 2 or 3 months at least.

Right.

And we're attending some of the AI conferences but we're not -- it won't be like the ODSC West where we hosted a panel.

It'll be more in an informal manner.

And next, we'll move on to Jeff Bernstein with Cowen.

Just quick questions around the rad-hard business.

I think you talked about also having a high-rel or kind of a rad-tolerant product, would be a different -- a lower price point and address some things like the [LEOS] or other more cost-sensitive market niches.

But can you just talk about what the plans are there?

And what the time line is there?

Sure.

So that requires a different die and it can -- that one could be offered in both plastics and ceramics.

With the rad-tolerant, we are actually at a position now where we've actually started qualification.

We're little bit ahead of the rad-hard as far as our internal qual.

As far as the customer engagements, we've already started those.

There are a few programs that we're addressing now and they are determining what direction they're going.

I will say the nearest potential is actually a program that sending a probe to Mars to do some mining on the planet.

And so that's certainly one that could be a short term -- or, I should say, our first win.

But certainly, there's been a significant amount of interest in something that's a little less robust than the rad-hard.

And at this time, there are no further questions.

Thank you all for joining us today.

We look forward to speaking with you again when we report our fourth quarter fiscal 2019 results.

Thank you.

Thank you.

That does conclude today's teleconference.

We do appreciate your participation.

You may now disconnect.