Vicor Corp (VICR) 2018 Q1 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Vicor Earnings Results for the First Quarter ended March 31, 2018, Conference Call, hosted by James Simms, and Dr. Patrizio Vinciarelli.

My name is Sam and I'm the event manager.

(Operator Instructions)

I would like to advise all parties this conference is being recorded for replay purposes.

I now would like to hand it to James.

Please go ahead.

Thank you, Sam.

Good afternoon, everyone, and welcome to Vicor Corporation's earnings call for the first quarter of 2018.

I'm Jamie Simms, Chief Financial Officer, and with me here in Andover, are Patrizio Vinciarelli, Chief Executive Officer; and Dick Nagel, Chief Accounting Officer.

Today, we issued a press release summarizing our financial results for the 3-month period ended March 31.

The press release is available on the Investor Relations page of our website vicorpower.com.

We also filed a Form 8-K earlier today with the SEC related to the issuance of this press release.

As always, I remind listeners this conference call is being recorded and is the copyrighted property of Vicor Corporations.

I also remind you various remarks we may make during this call may constitute forward-looking statements for purposes of the safe harbor provisions under the Private Securities Litigation Reform Act of 1995.

Except for historical information contained in this call, the matters discussed on this call, including any statements regarding current and planned products, current and potential customers, potential market opportunities, expected events and announcements as well as forecast sales growth, spending and profitability, are forward-looking statements involving risks and uncertainties.

In light of these risks and uncertainties, we can offer no assurance that any forward-looking statement will, in fact, prove to be correct.

Actual results may differ materially from those explicitly set forth in or implied by any of our remarks today.

The risks and uncertainties we face are discussed in item 1A of our 2017 Form 10-K, which we filed with the SEC on March 9, 2018.

Please note, the information provided during this conference call is accurate only as of today, Tuesday, April 24, 2018.

Vicor undertakes no obligation to update any statements, including forward-looking statements, made during this call.

And you should not rely upon such statements after the conclusion of this call.

A replay will be available beginning at midnight tonight through May 9, 2018.

The replay dial-in number is (888) 286-8010, followed by the pass code 74685691.

In addition, a webcast replay of today's call will be available shortly on the Investor Relations page of our website.

I'll start this afternoon's discussion with a review of our financial performance for the first quarter.

Dick will comment briefly on our first quarter implementation of ASC 606 and our outlook for income taxes.

And Patrizio will follow with comments about current business conditions, after which, he will take your questions.

Beginning with consolidated results as stated in this afternoon's press release, Vicor recorded total revenue for the fourth -- excuse me, for the first quarter of $65.3 million, representing a sequential quarterly increase of 11%.

This figure was 20% higher than revenue recorded for the first quarter of 2017.

Revenue associated with our advanced product portfolio rose 12% sequentially.

A comparison of first quarter 2018 to first quarter 2017, clearly illustrates the substantive shift underway for our revenue mix toward our advanced products, which rose 46% period-to-period from our legacy products, which rose 8% period-to-period.

Quarterly international revenue increased by 8.5% sequentially.

Turns volume, that is orders received and shipped within the quarter, totaled $15.4 million, representing approximately 24% of first quarter revenue.

Gross profit margin dollars increased at a faster pace than revenue dollars, 12% versus 11%, as total gross profit margin rose sequentially from 46.3 -- to -- excuse me, to 46.3% from 45.8%.

As stated in today's press release, gross profit margins are expected to improve, driven by higher volumes and economies of scale.

As reported for the fourth quarter of 2017, first quarter gross margins for VI Chip product lines, again, exceeded those for our legacy BBU product lines, representing further evidence of the scalability and inherent cost-effectiveness of our proprietary VI Chip manufacturing processes and packaging technology, which can support gross margins commensurate to levels achieved with advanced power management ASICs.

Our operating expenses for the first quarter were essentially unchanged with seasonal increases in audit and reporting fees offset by a reduced spending in R&D.

Pretax income totaled $4.1 million for the first quarter.

We recorded an income tax provision, reflecting state and foreign amounts of $134,000.

As such, net income for the first quarter was $3.9 million, representing $0.10 per diluted share compared to $0.04 per diluted share recorded for the fourth quarter of 2017.

Recall our fourth quarter EPS included approximately $0.02 per share of net income associated with the year-end change in accounting for alternative minimum tax credit carryforwards.

Our first quarter EPS includes no unusual or nonrecurring tax influences.

Turning to the balance sheet, DSOs were steady at 44 days, up from the prior quarter's 43.

Given the increase in sales, net receivables also increased rising $7.1 million for the quarter to $41.6 million.

Portfolio quality remains high.

Inventory also rose sequentially, increasing $2.5 million, largely a reflection of rising material and component purchases to meet our increased backlog.

Annualized inventory turns declined to 3.5 from 4.0 for the fourth quarter, reflecting higher than usual levels of safety stock, given industry-wide raw material lead times and other supply chain uncertainties.

Cash and cash equivalents, sequentially decreased $1.6 million for the first quarter ending at approximately $42.6 million.

This decline was due largely to a $7.8 million increase in working capital driven by higher sales.

This increase was offset by the $4 million of net income, $1.3 million realized through share purchases through our employee stock purchase plan and stock option plans, and capital expenditures that were sequentially lower by just over $500,000, totaling $1.6 million.

Our -- we anticipate operating cash flow to turn positive as the year progresses and intend to fund near-term capacity expansion from operating cash flow, and possibly, the sale of real estate asset.

To conclude my review of the first quarter, total employee headcount as of March 31 increased to 995 from 980, due to an increase in temporary staffing.

Total full-time employment was essentially unchanged.

As addressed last quarter, productivity continues to improve with level loading of quarterly production and longer-term visibility into our growing backlog.

Beginning with the first quarter, Vicor adopted ASC 606, which sets forth new guidance for how we recognize revenue.

Dick Nagel will now describe the impact of this adoption on our results for the first quarter and going forward.

Dick?

Thank you, Jamie.

As discussed during the last 2 quarterly earnings calls, on January 1, 2018, we adopted ASC 606.

Revenue from contracts with customers, which codifies the U.S. GAAP requirement for revenue recognition.

The adoption was addressed in our recent Form 10-K filing.

We utilized the modified retrospective method of adoption, which instead of restating prior year's data, led us to make a single adjustment of approximately $3.7 million to retained earnings as of January 1, 2018.

This increase in retained earnings represents the net effect of application of ASC 606 to existing customer contracts subject to such application.

The $3.7 million credit to retained earnings is reflected in our first quarter financial statements.

There was no cash impact from the adoption of the new guidance.

As I addressed last quarter, the most significant impact of ASC 606, going forward, is on the timing of recognition of sales to our stocking distributors.

As of January 1, we no longer defer revenue and the related cost of sales on shipments to stocking distributors.

With ASC 606, we record revenue at the time of sale to the stocking distributor, while also recording sales reserves, based on our historically based estimate of returns and allowances provided to the stocking distributors.

This is why Jamie no longer has to discuss stocking distributor activity in his remarks as such activity is no longer subject to different treatment.

Turning to our tax outlook for 2018, we are forecasting sustained taxable income for the entire year, which at some point, would cause the release of an allowance of approximately $33 million against our deferred tax asset, generating a significant onetime tax benefit.

With that, I'll return the call to Jamie.

Thanks, Dick.

Turning to our second quarter outlook, given recent increases in backlog, which totaled $90 million as of March 31, we anticipate a 10% sequential increase in consolidated revenue for the second quarter, with expanding gross margins and net profitability.

I must remind listeners, as I do each time I speak with you, our operating and financial forecast are subject to sudden and anticipated -- unanticipated changes.

So with that, I'll turn the call over to Patrizio.

Thank you, Jamie.

As reported, the first quarter was categorized by strong bookings from both customers for our legacy products and the expanding base of customers for our advanced products.

Factorized Power systems, the bus 48-volt to the point of load are gaining traction in demanding applications, including hyperscale data centers and supercomputing.

Notably, GPUs are pervasively being used both as accelerators of CPUs-based platforms and as dedicated compute engines, particularly for machine learning and related artificial intelligence applications.

Vicor is uniquely positioned in this high-growth market, with our Power-on-Package modular current multipliers, MCMs.

Enjoying design wins for the highest profile opportunities.

Aside from directly contributing to our revenue growth, starting the second half of 2018, success in converting advanced GPUs to 48-volt is accelerating the transition of data centers and automotive systems to a 48-volt infrastructure, that is central to Vicor's power system architecture and power components methodology.

Having to supply 48-volt GPUs within a 12-volt legacy infrastructure that is no longer capable of supporting higher bandwidth and connectivity, it's the CapEx, for change, that we will be pursuing.

As always, I could go on, but I'm sure listeners have specific questions, so I will open the call.

Operator?

(Operator Instructions) We do have a question.

The first question comes from [Don McKenna].

Patrizio, can -- based on the design wins that you've got now, can you kind of quantify the size of the addressable market that you see out there?

I don't know that I can quantify it in a meaningful way.

I think it's safe to say -- first of all, it would have to be done by segments.

And there are a number of different segments that are undergoing transitions that favor our -- the power components and power system architecture and technology.

If we focus, as an example, on the data center space, as you may know, Google played a pioneer role with respect to adopting 48-volt system in its data centers.

And now this goes back a number of years.

They contributed one of their systems to open compute to accelerate conversion of the industry at large to 48-volt in an effort to, in effect, expand an ecosystem that would be beneficial to all.

But frankly, it was not until recently, until leading companies, the GPU arena made the switch to 48-volt that within the data center space and the major players in that space, we started seeing a definite sense of urgency with respect to making a transition from the very short 12-volt infrastructure to 48-volt.

There are compelling reasons now to do it, given our pervasive GPUs are becoming within a variety of systems within the center space.

I expected that it's now a matter of relatively few years.

And this is not an overnight affair because of the complexity of this infrastructure.

But it's going to be a matter of years before we're going to see broad conversion.

And with that, the rest of the market, just in data center space, to answer your question with respect to one of the market opportunities, will expand greatly because Google, while being a very major player, is clearly not the only player in that space.

I could address your question in the context of other end markets.

In the long term, well, the data center space remains particularly interesting for us and ripe with opportunities.

It's really not the only major area of opportunity.

So I think in a nutshell, it's fair and safe to say that the addressable market is expanding very rapidly.

I'll make one more comment with respect to this.

Historically, particularly in data center space, Intel servers have been dominant.

Intel, historically, has relied on its own on-chip regulators to lessen the challenge for their customers in applying their CPUs.

But what we're seeing now more and more, and a progressions from opportunities are really coming up more and more frequently on a monthly basis.

We're seeing expanding requirements for what we call XPU solutions, AI, ASICs, other type of devices that unlike Intel, CPUs require a lower voltage, not 1.8 volt.

But typically, 0.8 volt, going down to 0.7 volt, going down to 0.6 volt across the rapidly approaching and further along exceeding 1,000 amperes.

In that type of a socket, there's nothing else that can provide an effective solution as recently demonstrated by recent design wins and recent adoptions.

So that's another power force at play, because as we all know, artificial intelligence chips that more and more are getting developed by the key players in the data center space and other places to serve their unique needs in a tailored manner that maximizes the efficiency of their applications.

They're becoming more and more part of the strategy of the core of major developments.

So it's the combination of variety of forces at play that are coming together to increase our market opportunity and the addressable market.

Sorry for the length here.

No, that's quite all right.

But if I could try it a different way, in previous conference calls, you've indicated that as far as you can see we're going to be going through these increased revenue periods.

And with the forecast for the upcoming quarter being 10% or so, is this a scenario that you would see sequential increases in that magnitude, for let's say, 3 to 5 years, as you see these outside markets developing?

I think we all know, around long enough to make very long-term forecast, right?

It's a vicious game.

But I think I can safely say, as I indicated in my earlier response, that there are forces at play.

They -- the conversion from 12-volt to 48-volt, the competition to CPUs from GPUs and ASICs that require very high currents or very low voltages.

These are all perfectly matched to our capabilities.

And I would say, rather uniquely matched to our capabilities.

So as far as my eye can see, I see expanding opportunities.

And frankly, we are in the [annual] position of having to pick which particular customers, which particular applications we want to pursue.

And we are being more selective with respect to the ones that represent -- that get us opportunity for us in the long term.

The next question comes from [Jim Bartlett].

Yes, along that same line, could you tell us how many ASIC, GPU customers where you're now shipping product?

And how many are up design wins that you may have in this space?

So what can you expect to say a year from now?

So I'm not going to give you specific numbers for a variety of reasons.

But I will say that there are several engagements and the number is growing.

And the inquiries with respect to our willingness to support new developments is -- that's increasing.

So I think we're engaged with a multiplicity of customers in a range of applications in sub-1 volt XPUs that range from 700 amps to up one extreme for a -- in array around 16,000 amperes.

So there's a variety of applications and the growing number of opportunities.

Again, when talking about this particular class opportunities, I want to remind us that will not, while being extremely exciting, what our power component methodology should be entirely about.

The remarkable thing is that with the same engines, with the same power component methodology, with the same packaging technology, all of which are very unique to us and highly proprietary and very heavily patented.

We can address different types of applications.

So for instance, we're also involved, at the other end of the spectrum, in front end requirements that take us from 3-phase AC lines or high-voltage passes to the 48-volt node, which is, in fact, the center of gravity for our power system methodology, the hub, around which our solutions revolve.

So the high-current ASICs are very exciting and clearly a great opportunity for, I believe, many years to come.

This is a market that is now expanding, which literally didn't exist, measured in terms of applications requiring others amperes as recently as a couple of years ago.

So it's -- and within a couple of years, I've seen requirements grow from a couple hundred amps to 1,000 amps, and as we suggested earlier, well beyond that.

And we don't see that subsiding to the contrary.

We see it with both the data center space and supercomputing and other types of applications.

We see it expanding with more and more players looking at AI trips, playing a key role going forward.

Autonomous driving, other kinds of applications, not necessarily within the data center.

Can you help me understand the NBM module that you introduced that, the availability and then how this may enable you to penetrate legacy data centers that are on 12-volt?

How quickly could this have an impact?

So we're looking to start shipping NBMs into applications in data center space where, to your point, there is an existing 12-volt infrastructure that cannot be changed overnight to 48-volt, but which gets challenged by leading GPUs running on a 48-volt bus.

So we want to see the adoption of these kinds of devices, making it easy, efficient, cost-effective to use them in a 12-volt infrastructure.

And that's what our NBM solution is all about.

It enables conversion from 12 to 48 -- and by the way, those enables conversion from 48 to 12 in a very cost-effective way with very high efficiency, big efficiencies of the other 98% plus and going high later this year -- going higher later this year.

It's very cost-effective.

It's very dense.

It essentially circumvents the challenge of making immediate decisions with respect to how to power 48-volt loads in a 12-volt system or 12-volt loads in a 48-volt system.

Let me talk a moment more about that because I think your question was aimed at how do we use NBMs to enable the set of customers that would like to use a 48-volt GPU in their 12-volts infrastructure?

The NBM, again, will convert 12-volt up to 48, so that the 48-volt GPU load can be readily powered.

But this is complementary play and that complementary play has to do with using that same NBM for particular purposes to convert 48 to 12.

And how does that play into the role transition of a 12-volt infrastructure to 48?

It plays by, in effect, taking away the anxiety that some customers might have with respect to making the transition.

Because it's a big bet with very profound ramifications.

So many of these customers have been long accustomed to their 12-volt multiphase regulators as the preferred way of powering CPUs and memory in their servers.

But they all recognize that there are significant efficiency benefits, infrastructure benefits to 48-volt.

So by having a device like an NBM that can convert 48 to 12, we're going to be facilitating a conversion of the infrastructure to 48.

So that the infrastructure can readily power 48-volt GPUs, it can power 12-volt systems, legacy systems through the NBMs as a stepping stone to eventually powering it all directly from 48.

So it is a key enabling building block that can facilitate use of 48-volt GPUs, in particular, 48-volt ASICs in a 12-volt infrastructure.

And conversely, can power legacy 12-volt loads in a 48-volt infrastructure.

That's facilitating the transition from 12 to 48.

And when do you think you'd start seeing some impact of this?

Well, so we are beginning to see some early requirements.

I think in terms of a meaningful impact on revenues, I don't anticipate that until 2019, where our programs that should start in the second half of '18.

But frankly, I think that in the very short term, the 48-volt GPUs, in particular, are going to be powered from within their system that doesn't require -- that is self-sufficient in terms of it being powered directly from the AC mains.

But I think over time, and again, that may take 6, 9 months, there should be some significant contribution to the revenues from the NBM opportunity.

We see it as a very exciting product.

Our NBM is a lot denser, more cost-effective than any competitive alternative.

And it would become even more so when we roll out our 4G, our next-generation technology later this year.

But even with our older generation technology, it is 3 or 4x smaller than any competitive alternative.

And it's more cost-effective than any competitive alternative.

So we think we have the winning card to play in that space.

And sort of a childish analogy, is this -- getting sort of the -- you get the nose under the tent with this and then -- but that may facilitate a broader conversion to 48-volt in data centers that are now really just 12-volt.

Yes.

I think the NBM is a building block that can facilitate that.

I think, again, what has recently been announced regarding [sale of the -- our GPUs] that reason itself is going to change the industry, I believe.

But facilitating a transition in every way we can is certainly part of our mission.

The next question comes from [Jon Andersen].

I had 2 questions if I could.

My first question relates to the tenor of orders in the current quarter.

You've just landed your -- perhaps your -- one of your biggest customers ever.

Is it logical to assume -- I think on your last call, you talked about orders sequentially expected to grow throughout the year in each quarter.

Is there visibility to expand upon on the call here for listeners?

Well, we've seen forecasts from key customers going up for the balance of this year.

In some case, it's going up by significant multipliers.

I think with some of these applications, the ramp in earnest will start in Q3.

But looking at Q2 thus far, we're doing very well.

We're well ahead of where we were at this time last quarter.

So the outlook in the near term in terms of actual bookings have already taken place.

And the outlook in terms of forecast and the demand that is forecasted by some of the key customers, they're all suggesting strength going forward.

And again, that fits with the general logic of the solutions and the applications and their respective opportunities in their marketplace.

Great.

And my other question relates to a note in your annual report where you referenced, you guys, in late 2017.

We know you've been working with Google for 5 years, that you're also kind of working with other very large hyperscale data center players.

Can you give any update there on -- we just saw Facebook last week talk about starting a new ASIC division.

Without naming any new potential customers, can you give any update on that initiative?

So with respect to customer names, the ones that are in public domain we can talk about; the ones that are not, we won't.

I'm not sure I fully understood the second half of your question.

If you could maybe rephrase it, in particular, regarding new ASIC developments.

I think there was just within the news last week, Facebook was rumored to basically be starting a new division focused on implementing ASIC chips just like Google has been doing.

So I just was referencing that just in relation to the note in your annual report.

Because I know you guys said in that note, you're working with multiple hyperscale data center players.

So just without specifically talking about Facebook, can you give any -- are there other opportunities that are kind of maybe nearing the Nvidia finish line that could be groundbreaking new customer relationships in the back half of the year?

So no comments regarding Facebook, particular, and no comments with respect to other key data center customers.

I can say that as suggested earlier, we're working to power a variety of high-current ASICs.

As you point out, Facebook recently announced or is looking to hire key people to staff a key development in this general area.

Well, guess what?

As we all know, they're not alone, right?

Fundamentally, advanced ASICs that are customized to process applications with unique advantages for these companies are going to be commonplace to all the key players.

They all have the resources to do it.

They all have the will to do it as far as we know, and they're doing it.

And to the extent again that they have, with these developments, a common denominator need for devices that ran on very low voltage nodes, 0.6, 0.7, 0.8 volts at very high currents, that's a proxy for a current multiplier.

And a Factorized Power System Architecture is the one that has got current multipliers that, at these very high currents, can lessen the challenge of delivering 1,000 amps by a factor of 50.

So fundamentally, we are the only ones with the technology to take a 1,000-ampere load and convert it to 20 amps or even less, 15 amps.

We are developing for 1 customer a so called [K-172], which is a 72:1 current multiplier.

So that will take 1,000-amp load down to 15 amps.

And that's all that, with our technology, you need to deliver to the ASIC package.

So your point is Facebook is going in that direction and in some other public record that they are, so are the other ones.

And guess what?

They're all in need of low-voltage, high-current solutions because, again, unlike Intel, with a long time and billions of dollars of development, R&D, to develop a chip, actual solution.

By the way, the Intel on-chip solution only reduces the challenge by roughly a factor of 2. In other words, if an Intel chip consumes 400 amps, it still needs to be fed with about 250 because the on-chip regulator can only convert 1.8 volts to, let's say, 0.8 volts.

And that's roughly a factor of 2 division in voltage and multiplication in current.

We can affect much, much larger current multiplication ratios.

Again, in the early engagement with PEZY in Japan, with their center customer in the U.S., we had -- our multipliers, I think they were in the 48:1 or 64:1.

We're now going to 72:1.

This provides, again, a huge advantage in terms of lessening the challenge of delivering the current, freeing up precious I/Os before -- for connectivity as opposed to power delivery.

So this is a very timely development that we think we are well positioned to address, not just with one customer but with a large multiplicity of them.

I'm going to cede the call to other shareholders.

Before I do that, I just want -- I'm sure I speak for other shareholders, and I'm relatively a new one with our firm, but I personally hold a lot of admiration for your forward vision in building this company.

You've been very tenacious in sticking to kind of predicting where the market was going to go.

And for that, my hats off to you, and thank you for all your efforts and really appreciate it.

The next question comes from [John Dillon].

Congratulations.

15% percent sequential bookings is just outstanding.

And Jamie, is the backlog of $90 million, is that a record for you guys?

I guess so.

I believe it is.

I can't remember ever having a $90 million backlog, so congratulations.

And Patrizio, I also wanted to thank you for the explanation on the currents and the amps because what I think I'm hearing is, is that Intel's got a built-in regulator, so they only need 1.8 volts.

But the other guys are going down to 0.9, 0.7, 0.6 even, which is going to need a lot more amps.

And this is where your technology is really, really shine.

So this is really your sweet spot, and this is the trend in the industry.

That's what I'm hearing.

Is that correct?

That's correct.

Another way of saying that is that, as you know, we've had some remarkable wins powering Intel processors.

In spite of the fact that they only need a couple of hundred amps at 1.8 volt.

Right.

It's not even your sweet spot, exactly.

And at that level, the 12-volt bus in a regulator, which doesn't really multiply current, it averages the voltage down.

And by averaging the voltage down, it can provide more current.

But fundamentally, and I don't want to get too technical here, but the analogy I like to use is, in effect, getting water at the right temperature by mixing the hot water faucet flow with the cold water faucet flow.

When you're at 1.8 volt and your power source is 12, 1.8 divided by 12 is about 15%.

And if you open the hot water faucet low and you draw most of the water from the cold water faucet, you can get down to 1.8 from 12 volt relatively easily.

But doing it at 0.6 volt makes the trickle of water out of the hot water faucet which, in technical jargon, is called duty cycle, much more challenging, not to mention the fact that power levels are going up and densities are becoming more challenging.

So to your point, being able to win at 1.8, powering an Intel processor, is indicative of the competitive advantage we have powering sub-1 volt ASICs, be they GPUs or other kinds of ASICs at lower voltages.

And obviously, this would be devices that wouldn't come out of Intel fabs, they will typically come out of TSMC.

And their voltage nodes are going down and down and down.

And with that, the costs are going up and up and up.

And the appetite for more compute capability is, as we all know, escalating.

So we think that we've had some good hits at 1.8, we are in a sweet spot below 1 volt and I don't see anything else that can really challenge us there.

That also kind of reminds me that if the leading guys are doing this, and they're freeing up pins and they're providing a more efficient GPU or XPU, the other guys are going to really have to follow in order to be competitive, aren't they?

Yes, I think we all know how these things go.

There are visionaries within companies that foster that mindset who are willing to take risks in order to gain competitive advantages for their own products.

They realize that playing the commodity game of being extra conservative isn't going to be a good long-term strategy.

And then there are other companies that limit risk-taking to core developments, and they tend to be a lot more cautious when it comes to, in particular, the power technology.

But to your point, that's becoming a riskier proposition because it's becoming more and more of an impediment to getting the fundamental task done.

So I think there has been reluctance in the past on the part of some to embrace radically different new technology from the likes of Vicor.

That reluctance is getting overcome, superseded by other considerations.

And I'll tell you one thing, if it isn't by way of current multipliers or MCDs and MCMs, we're getting there by way of NBMs or other power components or potentially front-end products.

So these buyers are breaking down.

And once we're in and we prove that we are very task-worthy as a supplier, in terms of capacity, pricing, reliability, we expect that the doors will open with respect to more sensitive applications.

Yes, I agree.

I think the NBM is just a brilliant move on your part.

But I do have another question that's more related to bookings.

And I'm just wondering sometimes companies will prime the pump when they start placing orders with you or another company.

They have a big order upfront, and so then your bookings might tend to tick down a little.

So I'm just wondering, for the next quarter, is it safe for us to assume another 10% sequential increase in bookings?

Or are we going to see it go down a little bit because they've already put their big orders in?

I don't know if I'm clear on this, but...

We haven't yet gotten any big orders with respect to the recent GPU design win.

That's going to be second half of 2018 event.

Now we've gotten some initial orders, and they're now negotiable but nothing that has really impacted the bookings pattern today.

So we could expect another 10% sequential increase in bookings with the...

I don't want to get pinned down on -- I think what you can legitimately expect is progress on the bookings front.

On the revenue front, I think it's easier to forecast revenue, and that's why we're sticking our neck out a little bit with respect to the revenue forecast for this quarter than it is to forecast bookings.

Because particularly with big projects, whether they come in or a large order comes in, the first 2 weeks of the quarter, the first 2 weeks of the next quarter, they can make easily a 10%, 15% difference with respect to the bookings of the quarter.

So I'm not going to stick my neck out with respect to bookings.

But I have said that, thus far, in this quarter, we are substantially ahead of where we were last quarter, and we'll let the quarter go by.

And it sounds like there was no real priming of the pump from some of the big guys, so...

No, there has been no priming of the pump with respect to -- none whatsoever with respect to new projects.

Now nothing in the first quarter with respect to the bookings that have been reported.

There has been some heightened activity of late, but that's part of the second quarter opportunity.

The next question comes from [Alan Hicks].

My question is on -- well, first of all, can you break out the revenues per segment VI Chip, Picor and BBU?

So I think in the prepared remarks, Jamie pointed out that the year-over-year growth is primarily due to the advanced products.

And you can take those to be VI Chip and Picor products.

That was a 40-some-percent.

We have been doing pretty well with legacy products, which are showing their resiliency.

But again, that's not where the growth, going forward, is going to come from.

And as discussed in recent quarterly calls, we're now at a point with the advanced products where they are closing a pivot with respect to our growth curve overall because historically, they were not critical mass.

They are now critical mass.

And as I look at the daily bookings report, I see the legacy system type of solutions doing okay but essentially being slightly up, fundamentally level over a long time scale, whereas the advanced solutions, MCDs, MCMs, VTMs, PRMs, BCMs, SIPs, these are the products that have been driving the growth of late.

And they're going to be driving the growth for the foreseeable future.

We do have major opportunities with front-end products, particularly the RFM.

We talked about that in the past that we've had a major engagement in Japan, which is going forward.

We're seeing a number of interested parties in that product.

So that's not a classic legacy product, but it is not the kind of point-of-load device that we've been talking about in this conference call.

It's very complementary to it because it gets customers from their power source, whatever that might be, to 48 volt on the way to the point of load.

But those are going to become significant, I believe, over the next 12 months as well.

So all the action is with respect to these new products.

To give you yet a bit of a different flavor, when it comes to advanced products, what we're shipping is primarily what we call our 2G and 3G technology.

You made reference to Picor, obviously, a key part of our effort with their ASICs, in particular the 2G ASICs and the 3G ASICs, we've been able to make a great deal of progress.

We're now very close to -- and in fact, we already have with one of the 2 key ASICs, 4G platforms that are working on the bench.

Very soon, we're going to have the complementary piece.

And before too long, starting late this year and into next year, we're going to see our 4G technology give us yet another level of capability.

And that builds on the packaging technology but takes the control system to a much higher level of integration and efficiency and flexibility.

So we're layering in these other developments, which are coming in, being contributed by portions of the enterprise, in particular, the Picor portion.

Alan, to answer your earlier question more specifically, you will have to wait roughly 2 weeks until we file our Q, we will have all of the segment reporting broken down.

Plus, we'll have the new format of our ASC 606 disclosures.

What I wanted to get at, I was reading through your 10-Q for last year and Picor is extremely profitable at about 33% operating margins.

BBU, looks like its operating income fell in half to about 3% or 4%.

And -- but VI Chip is still lose -- still lost last year about $11 million, although improved, I know it was improving throughout the year.

At what point do we get to profitability on -- what revenue level on the VI Chips?

VI Chip is there.

So I think we're getting to the 50% margin area.

I think as we get into the second half of this year, while it is true, to your point, that Picor SIPs have led the way with respect to very high margins, gross margins in the 70%, 80% area, VI Chip products have the potential getting to comparable levels.

It's all to do with scale and capacity utilization and absorbing the significant infrastructure cost of the unique packages, the packaging technology that underlies the VI Chip power component.

But all of these developments, when you look, let's say, at the GPU with 3 Golden Chips on it, those are all VI Chips.

They are -- 2 of them are current multipliers, one is a modular current driver.

Those are all devices that will bring about economies of scale and with that, significant margin opportunity.

I made this point in the past, I know it was difficult to make it and probably hard to believe when margins for VI Chips were zero or negative.

I think it's becoming a lot more credible today.

I think we've taken it from negative 20%, 30% to positive -- to close to 50%.

And I think we're going to take it quite a bit higher with rising volumes.

So overall, you can far surpass 50% gross margins?

I believe, we're going to be past 50% starting in Q3.

Okay.

And my next question is when do you expect to announce the RFM product?

When will it ship in volume?

So we've been very selective with respect to that.

We will ship some nonnegligible quantities of it but just to one customer.

And we're doing that cognizant of the unique challenges of front-end products of that kind -- of that power scale.

I mean, this is depending on where to supply an 8-, 10-, 12-kilowatt device, there are unique application requirements that go with it.

We have a few different flavors of the device under development.

We want to complete that development before making it available to other customers.

We have -- I'm reminded every other day that we have 3, 4 major customers that are very eager to evaluate it and potentially adopt it, and we see some major opportunities with it.

But I think it's very important to keep our powder dry with respect to some of these things and roll them out when they achieve the level of maturity they need to have.

That's more of a 2019 opportunity?

No, I think we are very close to more customers for that product.

I think it's 2018 event.

But I think in terms of it being a large-scale contributor -- as a scale contributor, it'd probably be late 2019, maybe even 2020.

And on the PEZY supercomputer, is that shipping yet?

Yes.

So their company had some issues that they have overcome those issues.

And we look forward to a long-term relationship as a key supplier.

They are pursuing opportunities with us in a variety of fronts.

And do you have other opportunities in supercomputers?

Oh, yes.

I think you name the supercomputer, and we're lucky to be in it.

And then last question, are you on track for increasing capacity?

Yes.

So with respect to that very key question, so as you've heard me say, we've been -- we're assessing a strategy, and we're continuing to reassess it.

In the October time frame of last year, we were very close to actually acquiring some space to develop but in 70,000, 80,000 square feet of incremental capacity capability.

We have decided since then to partner with a company that has certain complementary capabilities, and that's progressing very well.

We are pretty far along with that to bring about a capacity that will relieve some of our existing bottlenecks.

And that gives us the breathing room we want to have to make a judicious choice.

We're actively looking for space in the 200,000, 250,000 square feet area in -- nearby but far enough away from other facilities so that a meteorite could not hit both places at the same time which is, believe it or not, is also a concern for some potential customers.

When we actually pull the trigger on that, I do not yet know.

It could happen relatively soon.

But yet, we may be able to find other relief valves that enable us to expand the capacity capability we have within our Federal Street facility, which is about 250,000 square feet, so that we don't have to divert our focus from the ramp that we are in the midst of.

So we like to defer to the extent possible the acquisition of additional vertical integration capacity to the extent we can source it from the outside without any compromise, either in terms of quality, capacity or cost effectiveness.

And we think we can accomplish that combination of goals.

If there is one more question, we'll take it, otherwise...

Yes, you do have a couple of others, but the next one comes from [Ron Feinstein].

Patrizio, the biggest compliment I can give you is in 10 years I haven't asked any questions, so thank you for your results.

I just want to ask whether the foreign policy and the trade issues that could be coming for China or elsewhere, will that influence or affect your business at all?

In a very minor way, I think a few million dollars if a recent decision by the Department of Commerce holds for the next 7 years with one Chinese customers.

But otherwise -- so it's immaterial to our prospects, particularly in the near term.

I think the opportunity there was -- some are longer-term opportunity.

And I believe that that's an issue that is likely to get resolved before too long.

But in any case, it wouldn't be certainly affecting our prospects in the near term or the long term, so we're not particularly concerned about it.

And with that, we'll call it a day.

We'll talk to you in a few months, actually, at the shareholders' meeting in June.

Have a good day.

Thank you.

Ladies and gentlemen, that concludes your conference call for today.

You may now disconnect.

Thank you for joining, and have a good day.