Vicor Corp (VICR) 2018 Q2 法說會逐字稿

Good afternoon, ladies and gentlemen, thank you very much for joining today's Vicor Earnings Results for the Second Quarter ended June 30, 2018.

(Operator Instructions)

And then next, I'll introduce our host for today.

We have Mr. James -- Mr. James Simms, CFO; and Dr. Patrizio Vinciarelli, CEO.

Go ahead, please.

Thank you, Bart.

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

I'm Jamie Simms, CFO, and with me here in Andover are Patrizio Vinciarelli, CEO; and Dick Nagel, CAO.

Today, we issued a press release summarizing our financial results for the 3 and 6 month periods ended June 30.

This 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'll remind listeners this conference call is being recorded and is the copyrighted property of Vicor Corporation.

I also remind you, various remarks we 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 the 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 the light of these risks and uncertainties, we can offer no assurance that our forward-looking statements 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, July 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 the call.

A replay of the call will be available beginning at midnight tonight through August 8. The replay dial-in number is (888) 286-8010, followed by the passcode 68688022.

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

I'll start this afternoon's discussion with a review of our financial performance, 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 second quarter of $74.2 million.

Q2 represented a sequential quarterly increase of 13.7% from the $65.3 million recorded for Q1, and was 28.6% higher than revenue recorded for the second quarter of the prior year 2017.

Our consolidated gross margin rose to 48.4% for the second quarter, up from the prior quarter's 46.3% gross margin, and the Q2 2017 margin of 44.9%.

Operating expenses sequentially declined on a relative basis from 40.6% to 37.2% of revenue.

Operating income rose to 11.2% of revenue or $8.3 million for the second quarter.

For the quarter, our effective tax rate was 4.4%.

And we recorded a net provision of just $363,000, resulting in net income after a minority interest of $0.19 per diluted share as compared to $0.10 per share earnings last quarter.

The diluted share count used in the second quarter EPS calculation was 40,646,000, up from the prior quarter's diluted share count of 40,167,000.

Notable revenue events in Q2 for advanced products included a record volume of Picor SIP regulator shipments for 48-volt to point-of-load applications.

And initial shipments of MCMs and MCDs providing our Power-on-Package solution for a high-performance GPU application.

Quarterly international revenue increased by 20.3% sequentially.

Turns volume, that is orders received and shipped within the quarter totaled $13.5 million, representing approximately 18% of second quarter revenue, reflecting a continued increase in our consolidated backlog, which exceeded $103 million at the end of Q2.

Product gross margins benefited from the efficiencies of higher production volumes.

But were negatively impacted by higher material costs.

We are carefully monitoring material costs, given supply chain constraints, which may continue into 2019 for commodities such as ceramic capacitors and discrete semiconductors.

Nevertheless, we have not changed our expectation for sequentially higher gross margin percentages through 2018, driven by higher volumes and economies of scale that will further reduce average unit cost.

Operating expenses for the second quarter rose 3.9%, sequentially, in absolute terms, after being essentially unchanged from the fourth quarter to the first quarter.

The impact of our annual merit increases for salaries and wages as headcount is our largest expense, an increase of sales commissions paid associated with an increase in commissionable sales, and certain one-time charges associated with severance accounted for the bulk of the increase in operating expenses.

I'll return to these charges in a moment.

Our largest operating expense category, R&D has been steady on an absolute basis since Q2 2017, the height of the development of Power-on-Package technology.

For Q2 2018, R&D expenses represented 15.4% of consolidated revenue.

While we believe, 15% of revenues for engineering expense is an appropriate level of R&D spending for Vicor, we also believe having invested upwards of $400 million and secured a comprehensive patent portfolio over the last 10 years, our R&D spending can expand at a lower rate than our expected revenue growth rate for the foreseeable future.

As such, we expect the yearly rate of increase in R&D spending to be in the single digits.

Marketing and sales activities currently represent our second largest expense category at 14.3% of revenue, and it increased as we have built out the global sales and field applications personnel and infrastructure needed to execute our strategic transition.

While mindful of the cost of lengthy, evangelical efforts needed to promote Vicor's highly differentiated products, we believe the company has crossed the chasm, so to speak, given the momentum achieved in the 48-volt to point-of-load market segment we have created.

As such, we believe growth in marketing and sales spending should lag top line growth as our customer base expands beyond early adopters.

Sales commissions, customer support activities and T&E will likely grow at rates commensurate with our sales growth.

But total payrolls should increase at a lower rate.

Accordingly, we expect the ratio of marketing and sales expenses to revenue to decline.

G&A expenses represented 7% of revenue in Q2, and are expected to decline as a percentage of annual revenue, but of course, will vary quarter-to-quarter depending upon the timing of audit and filing expenses.

Returning to the one-time charges incurred during the quarter, we are closing one of our custom subsidiaries and transferring customer engagements to other subsidiaries, effective 12/31/2018, as a part of our ongoing initiative to streamline operations and improve our cost structure.

To cover this closure and other severance and personnel related cost, we incurred a total of approximately $500,000 in charges for the second quarter.

Other income swung to a negative value as we recognized a $300,000 foreign currency valuation loss, largely related to the weakening of the euro in April and May.

While substantial currency moves do occur they are difficult to predict and our expectation for other income remains positive going forward.

Pretax income totaled $8.3 million for the second quarter.

As stated, we recorded an income tax provision, reflecting federal state and foreign amounts of $363,000 for the quarter.

I'll now take a moment to address our evolving tax position and to clarify the accounting for our deferred tax assets.

As of December 31, 2017, the balance of the 100% allowance against the value of Vicor's domestic net deferred tax assets stood at $33 million.

This allowance was established over the years to reflect the likelihood considered more likely than not at the time, Vicor would not return to a sustained level of profitability that would allow the company to utilize its deferred tax assets, which represent the cumulative value of deductible temporary differences, tax credits and tax loss carryforwards to offset future taxes.

Due to Vicor's improving financial results, over the coming quarters, we will be assessing the need to continue to maintain this valuation allowance based on the likelihood of a sustained -- sustainable level of quarterly profitability, allowing the full utilization of our domestic DTAs.

Elements of this assessment will include our ability to accurately forecast taxable income, the number of quarters of successive profitability appropriate to support a decision to release, i.e, reduce the allowance, and the amount of the allowance to be released.

If and when, we determine the valuation allowance should be released, we would debit the balance of the allowance at the time and credit income tax expense by the same amount, resulting in a potentially substantial tax benefit, in turn resulting in a potentially substantial increase in reported net income for the quarter in which the release occurs.

Please note, this release transaction would have no direct impact on cash flow.

Also note, due to our recent profitability, the company has been utilizing available net operating loss carryforwards and tax credits to offset taxes due on taxable income based on our estimated total tax provision for the year.

As such, the balance of DTAs and the valuation allowance against those DTAs have declined.

But because the DTA balance is reported on our quarterly balance sheets on a net basis, reflecting the allowance, and because we prepare our quarterly tax expense calculations based on full tax-year assumptions, in other words, we don't adjust or true up the valuation allowance for the financial statement footnotes in our Form 10-Q filings, the updated balance of DTAs and the balance of the allowance against them may not easily be determined by investors until we file our 2018 Form 10-K sometime in early March 2019.

Assuming continued profitability, if and when, we were to decide to release the then current valuation allowance, the amount of such a release would be lower than the $33 million balance recorded as of December 31, 2017.

However, at the present time, we cannot reasonably estimate the timing of the potential release, the amount of the allowance to be released or the of balance the DTA's at the time of the release.

We are bringing this to your attention at this time in consideration of Vicor's improved profitability, driven by what we considered to be sustainable, company-specific factors.

While there is no number of profitable quarters, the accounting profession and the SEC consider a minimum threshold for supporting a decision to release an allowance, we believe the company is well-positioned to sustain the recent trend of improving performance.

As such, although, we have no schedule for doing so, we believe it is more likely than not we will release some portion, if not all, of the then current allowance within the next 4 quarters.

Turning to the balance sheet.

Cash and cash equivalents sequentially increased $11.2 million for the second quarter, ending at approximately $53.9 million.

This increase reflects operating cash flow of $9.3 million and $3.6 million of proceeds from the exercise of employee stock options during the quarter, offset by capital expenditures of $1.7 million.

Note, just over half of the proceeds from stock option exercises were associated with the May merger of our Picor subsidiary into Vicor, which while -- I will address at the conclusion of my remarks.

Given the increase in sales, net trade receivables also increased, sequentially rising $3.4 million or 8.3%, ending the quarter at $43.9 million.

DSOs rose slightly to 46 days, up from the prior quarter's 44 days.

Portfolio quality remains high.

Inventory, net of reserves also increased sequentially, rising $2.8 million or 7.2%, largely reflecting rising material and component purchases to meet our increasing backlog.

Annualized inventory turns were steady at 3.5 quarter-to-quarter.

Winding up my review of the second quarter, total employee headcount as of June 30, increased to 1,024 from 995 due to increased temporary and co-op staffing.

Total full-time employment was essentially unchanged, up 3, from 969 to 972.

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

My final comment is on the merger as of May 30 of Picor Corporation with and into the parent, Vicor Corporation.

As a result of the merger, the separate corporate existence of Picor ceased, although its operation remain a business unit within Vicor.

And we continue to present Vicor as a distinct segment in our publicly filed financial statements.

Both Picor and the VI Chip subsidiary were established as separate corporations in order to facilitate an independent status at some future date, reflecting their distinct operational and product characteristics.

Over time, employees were awarded options for the purchase of subsidiary stock, and some options were exercised, resulting in the ownership of subsidiary stock by employees and retirees.

Approximately 3 years ago, the Vicor board authorized the development of a plan that would provide liquidity for holders of subsidiary options and stock as independent status was less likely for both subsidiaries.

We completed the Picor merger in the second quarter.

To perfect the merger, holders of Picor common stock and Picor stock options received an equivalent value of Picor -- excuse me, of Vicor common stock and Vicor stock options, respectively.

And the Picor Corporation amended and restated 2001 stock option plan and any options outstanding thereunder were assumed by Vicor.

The merger and the option plan assumption did not represent a tax event for either employees or the company.

Similarly, there was no impact of the merger on Vicor's consolidated financial statements or any impact on our segment reporting for the 3 and 6 months ended June 30.

The primary impact of the merger and option plan assumption was an increase of approximately 500,000 shares in Vicor's fully diluted share count calculation.

Management plans to address liquidity for holders of VI Chip options and shares, possibly with the same methodology recently utilized for Picor holders.

We anticipate being able to complete the transaction, involving our VI Chip subsidiary in 2019.

Our expectation is the completed transaction or transactions will have no material impact on Vicor's consolidated financial statements.

Finally, turning to our third quarter outlook, given our increased backlog and visibility into customer requirements, we are forecasting a sequential increase in consolidated revenue with improved profitability.

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

As discussed, the supply chain uncertainties represent near-term risk for us as well as our customers, notably CM's experiencing supply constraints or the cost of potential tariffs.

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

Thank you.

As reported in our press release and addressed by Jamie, the second quarter of 2018 was characterized by substantial increase in EPS, net of higher material costs and one-time charges.

Total bookings increased sequentially to $87.5 million, the 10th consecutive quarterly rise, reflecting the long anticipated transition from 12-volt to 48-volt in data centers.

We believe that before too long, a similar transition will gain momentum in automotive.

Of the point-of-load, Vicor is uniquely positioned to enable high performance CPUs, GPUs and artificial intelligence assets for applications, including cloud computing, autonomous driving, 5G mobility and robots.

It is worth noting that power computing, artificial intelligence ASICs built on a 7 nanometer node, and consuming as much as 1,000 amperes cannot be efficiently powered without VTMs or MCMs, using Vicor Power-on-Package technology.

And the 1,000 amperes is what it takes to expand computing beyond the limits of Moore's law.

With the transition from 12 to 48-volt, Vicor is also uniquely positioned to provide high-density front ends, supplying 48-volt hubs from AC sources, doubling our market opportunity while enabling Vicor customers to leverage complete, high-density power systems from the source to the point of load to 48-volt.

To meet rising demand, we're expanding capacity, initially in our 250,000 square feet Andover factory, and in approximately 18 months with a second facility, specifically design to manufacture proprietary chips as in convertor housing package, the package technology needed to enable Power-on-Package.

As always, I'd like to limit my prepared remarks to a minimum as I would rather answer your pertinent questions.

So I will open the call.

Operator?

(Operator Instructions) Our first question coming from Ken Farsalas from Oberweis Asset Management.

Just me -- I'll start with 2 questions, and then hop back in the queue and give others a chance.

But can you be more specific in your capacity expansion plans going forward?

I know last quarter you spoke a little bit about the potential for using third party to help you with on the manufacturing side.

Can you just expound a little bit on what your plans are at this point to increase your capacity going forward?

Yes.

So as discussed in earlier calls, we have lined up third parties to facilitate certain process steps that can be taken out of our manufacturing lines in Andover to both align ourselves with certain core competencies and facilitate better capacity with the Andover factory.

With this initiative, we believe that the Andover factory can support upwards of $0.5 billion in sales.

We are in parallel pursuing opportunities for expansion in our general area, likely New Hampshire but possibly Massachusetts, we're looking at a variety of opportunities.

And we'll be closing in on one of these opportunities in the near future anticipating that the second factory will want to be up and running in approximately 18 months.

Okay.

And were there any 10% customers in the quarter?

And if there were, was anyone new to that list in the current quarter?

There was no 10% customer in this past quarter.

Our next question coming from Matt Vigneau from Shay.

And my question is related to the previous one.

But noting that the company has achieved the best gross margin in recent years and the continued progress expected in sales, I'd like to ask about, A, how much more efficiency can be achieved out of the existing facilities in Andover from a gross margin standpoint?

And then part B would be, as you contemplate the additional facility 18 months out, when this comes online, would it be reasonable to expect a step back in gross margins or would the additional capacity be possible in a format that maintains your gross margin progress?

Thank you.

So first with respect to expanding capacity within our existing facility.

As you can imagine, or as implied by discussions that have taken place in the past, we've had a large and largely fixed cost structure associated with -- in particularly in this area, manufacturing overhead to facilitate the unique processes and capabilities we have developed in our factory.

These resources will not have to scale up anywhere close to direct proportion to the increases in revenues that can be supported from the Andover factory.

In other words, as we get to the roughly $300 million level of the most recent quarter to $0.5 billion level, I do not see the bulk of our manufacturing infrastructure growing in cost by more than, I would say, high single digits.

So with that, we anticipate significant improvements in margins.

We also expect to be able to achieve reduced material costs associated with larger volumes.

So all this should support an expansion in margins.

I should note though that with respect to the material cost component, I suggested in the prepared remarks and as noted also by Jamie, particularly with respect to certain commodities in recent quarters, they behave really as commodities because of the demand levels with respect to those components.

But this is an issue that will settle out.

And again, looking at the 10,000 feet, the material cost structure of the products will be going down to some degree with increased volumes.

Now with respect to second half of your question, to your point, as we begin to invest in a new facility, we're going to have to start depreciating that facility, the equipment that goes in it.

And we're talking tens of millions of dollars' worth of equipment.

However, as suggested in earlier calls, we are first taking the steps necessary to expand capacity within the existing facility to increase the margins well above the recently reported 48% before we begin to have to carry depreciation cost associated with a new facility.

Also, in regards to the new facility, as suggested in the prepared remarks, this is going to be a facility that will be designed -- is being designed from the ground up to make the manufacturing of chips, our unique packages that are going to support initially all of the products in the most efficient manner, more efficiently than with the existing facility.

So in summary, there's going to be continuing improvements over the next several quarters.

At some point, there's going to be negative contributions arising from additional depreciation.

But those are going to be absorbed by the higher revenues and ultimately are going to be benefiting from the great efficiency of a manufacturing line that is architected to make best use of resources in making chips, something that when the Andover facility was originally designed wasn't even a figment of anybody's imagination.

Our next question coming from John Henderson from Kershner Trading Group.

One question regarding the cadence of orders.

On your last call, I think you had provided investors with the expectation that orders would grow sequentially throughout the year.

I was just checking in on that front.

Are the expectations still the same?

Yes.

The expectations are still the same.

Thus far, this quarter, we are ahead of the prior quarter, about the same rate as last quarter.

Understood.

And looking further ahead, is there any type of acceleration that you can talk about for Q4 for next year?

Is it just too early yet?

I think it's too early that there's a lot of moving pieces.

I suggested in prepared remarks that within the last several months, we've seen a sea change having to do with the many initiatives on the AI front.

The realization that many have come to make regarding the fact that when you need to power an ASIC with 400, 500, 600 amps of current consumption at 0.8-volt or 0.7-volt, the (inaudible) solution going through a 12-volt bus and a large multiplicity of buck regulators no longer works.

The expression the buck stops here comes to mind or the many bucks stop there.

They stop at a few hundred amps, practically speaking.

So we're seeing a lot more opportunity developing on that general front.

Also suggest in the prepared remarks, we're also seeing more opportunity taking customers from their AC sources to the 48-volt hub.

We already had one such engagement in Japan last year.

I believe that this is a precursor to future engagements of a similar kind where we're going to be providing the complete solution from the AC source to the point of load through Factorized Power building blocks all the way.

So all of these things will contribute to near term and longer term growth opportunity as far as I can see.

And final question.

I know when I attended the investor meeting for your annual report, you referenced Waymo as a customer.

And in Google's conference call, they referenced how autonomous vehicles usage was 8 million miles.

And it seems as though the opportunity has nearly begun.

Are you still of the opinion that it's going to be 2022 until we see full-fledged adoption there?

Yes.

We believe from a variety of sources that the promise of Level 4 and eventually Level 5 autonomous driving is coming to fruition.

I think it is going to start happening with the fleets for particular purposes in that general time frame.

We're very excited to be engaged in the best applications of this kind with, I think, companies that have had great leadership in that general field.

So this is a first step in our entry into the automotive space where we believe we have considerable opportunities going beyond the autonomous driving opportunity.

GPUs are another area of opportunity in that general space.

Next question coming from Mark Lanier from Pegasus Capital.

One of my questions has been answered.

The other one has to do with the subject of tariffs.

I understand the full range of uncertainty that surrounds this issue to a degree, but I wonder because you mentioned it previously in the call, what thoughts you have about the levels of exposure you may have for potential tariffs that people may be predicting?

That would be helpful.

Well, so it's obviously a very fluid situation.

I have to predict what is going to happen.

Thus far, we have not incurred significant incremental costs due to any tariff.

We do not know whether there will be any real impact from any of the tariffs.

Obviously, there's been talk of potentially having tariffs applied to all of the Chinese imports, in particular.

And we do source from China certain components.

But I will say on a general level that throughout the electronics supply chain, tariffs are going to have to be absorbed.

And to the extent that they stick, passed along to customers.

Obviously, if it is a de minimis impact, we're going to absorb them for goodwill.

If it were to become more significant, we'd have to reflect them in the price of the product.

I hope that answers your question.

Our next question coming from John Dillon from D&B Capital.

Hey, guys, congratulations, really great quarter, in particular, the backlog numbers.

I think a record, all-time record.

And also, it's really nice to see you generate that much cash, really, really nice job.

Thank you.

So my question is on the bookings.

I want to just see if I could get a little more color on the bookings.

With the NBMs and the Nvidia opportunities hitting in the fourth quarter, I think Nvidia's announced they're going to be shipping production quantities of their new systems in the fourth quarter.

Can we expect the bookings to have a sequential increase this quarter?

Yes.

As I mentioned earlier in answer to an earlier question, through July 24, we are substantially ahead of where we were at this time last quarter.

Okay.

Good.

Could we expect a 10% increase, do you think, on that?

I'm not going to make a 2-digit accurate prediction with respect to that.

Let's wait and see what happens.

But I think we are continuing to see improvements on the demand side.

Have you had many bookings for the Nvidia stuff yet or is it still coming this quarter?

Did you have...

We're not mentioning customer names.

But generally speaking, we are seeing bookings from all our new programs, including ones that have been made public.

Great.

And you mentioned supply chain constraints.

Do you feel that, that's going to slow down your revenue?

I mean, I heard from Jamie, you're expecting a sequential increase in revenue, but I didn't hear any kind of, is it going to be substantial or just a marginal increase in revenue?

We're going to see, I think, a substantial increase in revenue.

I think availability of certain components, in particular, semi caps and power semiconductors has been an issue.

There have been some price increases in that general area.

We absorbed those.

We incurred some incremental costs in Q2 because of those.

But we've been able to get the components we need in order to take care of our customers.

Okay.

Great.

So you have the parts.

You're just going to pay a little bit more for them?

Yes.

Yes.

And one last thing.

You said you're extending the performance beyond Moore's Law with the new 7-nanometer.

Can you kind of explain that a little bit more?

Is it because the way you deliver power, you're enabling these artificial intelligence systems or how does that do that?

Yes, so typically, we're not in the business of making the CPUs or the GPUs, far from it.

Our business opportunity is limited to enabling solutions, best-in-class solutions, which notably require power.

And the point there is that, it has become very clear that number one, Moore's Law has had a long run and a great run.

Intel was for a very long time able to keep raising the bar on performance.

I think it's no mystery that Intel CPU has not in recent times made progress at the rate of earlier times.

So Moore's Law has been, if not hitting the wall, has been coming into a different area of maturity and slowdown.

On the other hand, Nvidia as the prime example with its GPUs has been able to accelerate, no pun intended, performance of ASICs that are developed with a different methodology and on processes that require power delivery at the level of several hundred amperes and voltage nodes that are well below 1 volt.

So that's fundamentally different from the Intel paradigm.

So the Intel paradigm has been predicated on a different computing architecture and on an internal regulator methodology that sources power, typically a 1.8-volt.

So what you have in the computing field is, again, the Intel paradigm, which continues to leverage Moore's Law based on 1.8-volt feed to the CPU at typically 100 or 200 amperes on the one hand.

And then you have GPUs and other kinds of AI ASICs that are operating on the 7-nanometer node, typically TSMC in Taiwan.

And from those platforms, they can achieve very high level of parallel computing performance, but at the cost of having to feed those devices with a 0.8-, 0.7-volt to 400, 500, 600 amperes and rising.

It's those kinds of computing platforms that are going beyond the limits of Moore's Law.

And it's those kinds of computing platforms that with our chips and with our Factorized Power solutions, with our VTMs and even more so with our MCMs, with our Power-on-Package technology, we can best support.

Next one, we have Ken Farsalas from Oberweis Asset Management.

Just 2 more questions quickly.

The new facility, you talk about an 18-month time frame to have that facility up and running.

Is that being driven by a specific customer demand forecast that your customers have discussed with you?

In other words, what visibility do you have over that time frame for the need to have that type of revenue capacity?

And what do you think the CapEX would be for the new facility roughly?

So the math is roughly the same, all right.

We are growing at 30%, 35% annual rate.

And in the most recent quarter, we were an annualized at $300 million rate.

So just a linear extrapolation, I don't want to go out on a limb here, but I'm answering your question, in effect, at a high level.

In 12 months, with a 30%, 35% growth, the $300 million turns into $400 million.

And 6 months after that, we get pretty close to the capacity limits of Andover facility.

And obviously, we would not want to be that close to the capacity limits of the facility.

So we want to get a new facility up and running ahead of an actual need.

Now having said that, the elasticity with respect -- some level of elasticity with respect to the expected capacity of our existing Andover factory, we've been creative in the past with respect to figuring out ways on how to get more capacity out of a certain space.

And I'm sure there's still opportunity for some further expansion in that capacity should the need arise.

So this is roughly speaking the math, the sense behind the need to have another facility in place in approximately 18 months.

Whether it's going to be 13, 14 months or a little longer than 18 months, it's going to depend on how much more capacity we can get out of it.

And what actually happens with respect to demand from particularly large customers and customers are larger.

We have a more and growing statistical base of customers.

So that's a general rise in the tide.

And within that, there are some very large customers.

Even though we haven't had one in the last quarter with more than 10%, we may well have in the next year greater than 10% customers, maybe one, maybe 2. So their unique demands will also drive the decision point with respect to the additional capacity.

Okay.

Great.

And Jamie, just more of a housekeeping question.

Do you talk about the revenue split between advanced products and legacy products in the quarter?

We did not in our remarks, but I believe we do in the filing.

Okay.

Good point.

Generally speaking, I would say that both -- we look at the business as a power component business at the point of load and a power system business in the front end.

The power system business includes within it the legacy business of the classic Vicor bricks.

Both businesses are growing.

The growth rate of the power component business, that is the point of load business, has been significantly higher than the power system business.

But I suggested earlier, I would expect before too long the power system business growth rate will increase.

And long term, both business units should grow at comparable rates.

Next question coming from Wally Walker from Hana Road Capital.

As your backlog has grown as rapidly and significantly as it has, how should we think about the trend in converting that into revenue?

And was Q2 a reasonable template for doing so?

Yes.

I think that as suggested in the prepared remarks, we see the quarter we're in as a quarter in which we're going to be growing the top line.

We're not spelling out exactly how much because of a variety of factors.

But we see growth that is going to bring about further expansion in the margins.

We expect to cross the 50% level on gross margins based on the top line increasing to a level that will bring about greater economies of scale.

Next question coming from Alan Hicks from Ainsley Capital Management.

Can you give us an update on the RFM product?

So we have begun to branch it out into particular models for a broadening range of customer opportunities.

So the initial engagement, as you might recall, was with a Japanese customer that I referenced earlier, where that was the first.

We provided a complete solution from what in Japan is a 200-volt, three-phase AC source all the way to the point of load by way of large arrays, a multi-megawatt insulation, a large array of RFMs in a liquid cool system delivering to a 48-volt hub the power required to drive MCD and MCMs, powering a large array of CPUs.

That initial system is paving the way for new engagements with a number of customers.

They will not use the same 200-volt RFM because that's kind of specific to the Japanese AC mains.

What we'll need for our engagements are versions of the RFM that enable conversion from 400-volt and 480-volt AC sources.

Those are coming together.

We're having customer demonstrations in the next few months, again, setting the stage for RFM sales and complete solutions from the AC source all the way to the point of load.

So the initial product in Japan that was sort of -- was like a custom application for them or was it a central product you have?

Well, it was a solution, again, for the Japanese AC source, three-phase source which happens to be 200-volt to apply RFMs into data centers and other applications on a global scale.

It takes a variation on the initial RFM.

It's fundamentally the same product.

It's just minor alterations that enable it to operate from 400 or 480-volt sources.

So we've completed those improvements and those alterations and are engaging now in -- or beginning to engage in demonstrations with a number of interested customers.

So there is great deal of interest in this product for a number of reasons.

One is obviously the 48-volt hub taking hold in certain end markets and needing to supply the 48-volt hub from a variety of sources, including in particular three-phase AC sources.

Another factor at play is liquid cooling.

The RFM is uniquely adept at being cooled either in fine air, in an immersion bath or with more conventional water cooling, which conventional AC-DC front ends are not nearly as capable of doing.

In other words, we can support much, much higher density in the front-end than conventional AC power systems can because of the unique attributes of the RFMs.

As you might have read, liquid cooling is coming of age because of the drive to higher density in racks and generally speaking, in a variety of systems.

So all these factors converge to the merit of a solution.

It's architected in all these key elements from the RFM performing conversion from the power source to the 48-volt hub, followed by, in particular, Power-on-Package, MCD-MCM systems, taking power from the 48-volt hub and delivering it to the point of load.

So would you say you're standardizing some products based on that original application with PEZY?

Yes.

Yes.

We have a standard product now that can be configured to convert power from either 400-volt, 200-volt or 480-volt to a 48-volt or a 54-volt hub.

Okay.

So how soon do you think that would be shipping in volume?

Well, so I think we are engaging with customers that have a strong interest.

I think to manage expectations in this regard, I should point out that we are dealing with a product that is not a consumer product.

It's got a gestation period in terms of designing cycle that gets measured in -- as a ballpark, 12 months.

I think best case, it can be a little shorter than that.

In some case, it can be 1.5x or 2x that depending on the nature of the customer and the application.

But I think it is something that as we look out 1 year to 2 years will begin to make a substantial contribution to our revenues.

And by the way, the Japanese customer will be buying for 2019 substantial new quantities of our RFMs as well.

So it's not that we get a dry spell from here until broader shipments of RFMs.

Have you shipped, I think if I remember right, it was $8 million order from a Japanese customer?

Yes, it was very substantial at the time.

Has that all been shipped?

Yes.

So that's not included in backlog then?

No.

That was shipped last year.

Oh, last year.

Yes.

That was last year.

It hasn't contributed in recent quarters.

But those systems with that particular customer will come back in significant volume in the first half of 2019.

They had another larger system they were developing, is that...

A larger scale system, yes.

Yes.

Okay.

Then, it was maybe a year or 2 ago, I think Jamie mentioned you might be having a revenue line called power systems at some point.

And I would assume with the products like the RFM...

Yes, they fall in that category, absolutely.

So the RFM is the prime example of a power system product.

It's not the only one.

So the NBM is also a power system product.

So generally speaking, power system products are unlike point of load solutions, they are devices, converters that take a power source and convert it to some hub, be it 48-volt, 12-volt, as the case may be, to -- and these are in effect intermediate voltages to the point of load.

You can think of it as a stepping stone.

Fundamentally, in our kind of power conversion, the power comes from a source, typically, a high-voltage bus or an AC source, single-phase or three-phase and to get to the point of load, to get, let's say, to an AI ASIC, ideally, it would take a first step from the source to 48-volt and sometimes that's 54-volt.

And then from 54-volt or 48-volt, as the case may be, directly to the point of load.

The hub is, if you will, an analogy of an airline having airplanes that take customers from any initial airport to a variety of destinations by way of a hub.

A voltage hub plays a similar role.

We take the power from a variety of sources, take it through 48-volt on its way to the point of load.

And the power system business is to do with taking that first step.

I think you said at one point, the power systems part of the business could be 50% of all your revenues?

Yes.

Because fundamentally, it's the same amount of power, right?

Power is essentially conserved.

So these systems are typically -- have an efficiency in the 90s.

So whatever is consumed, let's say, by ASICs gets supplied.

If it gets supplied by a Factorized Power solution, comprising PRMs, VTMs or MCDs, MCMs, that same power comes out of the 48-volt hub.

It needs to be delivered.

It needs to be replenished at the 48-volt hub by front end systems.

So what you're dealing with is the same power delivery through this series of 2 steps.

And the value proposition with each of the 2 steps in terms of cents per watt is fundamentally in the same ballpark.

So that's the simple math that stands behind the factor of 2.

Okay.

One last question.

I was looking at the last 10-Q, and it was about the eliminations.

So I think you did $8.2 million, I think, in Picor revenues, but it gets reported as $4.5 million because part of it goes into the BBU revenue line.

Is that how that works?

Yes.

Well, so what is going on with that is that part of the very important role that Picor has played and continues to play within Vicor at large, it is the source of all our ASICs.

So we don't make typically AI ASICs, but we do design, develop and get fabricated a variety of Vicor unique controllers.

These are ASICs that Vicor has designed, developed and provided to Vicor as a whole.

And so there have been intercompany sales having to do with those ASICs.

Next question coming from Jim Bartlett from Bartlett Investors.

Real quickly.

On the new facility, assuming 250,000 square feet, what would be the capital cost?

I'm not going to spell out a number, but it is several tens of millions of dollars all-in.

It's not going to be all-in on Day 1. But obviously, the space itself and then a considerable amount of equipment going in.

Is it just going to be 250,000 square feet or could it be more?

Well, so we're looking at options that may include more than the land required to support 250,000 square feet, but I'm not sure that we're going to adapt on a campus that can support that.

So there's a lot of factors at play.

And this will all get sorted out in the next few months.

With that, thank you, and we'll be talking to you in 3 months.

Thank you, ladies and gentlemen.

You may now disconnect.