Peraso Inc (PRSO) 2016 Q1 法說會逐字稿

Good morning, and welcome to the MoSys first-quarter 2016 financial-results conference call. At this time, all participants are in a listen-only mode. At the completion of today's conference call, instructions will be given for the question-and-answer session. (Operator Instructions) As a reminder, this conference call is being recorded today, Friday, May 6, 2016.

I would now like to turn the call over to Beverly Twing of Shelton Group, investor relations. Please, Beverly, go ahead.

Thank you. Good morning, everyone. Joining me today on today's call are Len Perham, MoSys's President and Chief Executive Officer; and Jim Sullivan, Chief Financial Officer.

Before we begin the call, I would like to remind everyone this conference call will contain forward-looking statements based on certain assumptions and expectations of future events that are subject to risks and uncertainties.

Such statements are made in reliance upon the Safe Harbor provisions of Section 27A of the Securities Act of 1933 and section 21E of the Securities Exchange Act of 1934, which include, but are not limited to, benefits and performance expected from use of the Company's ICs and embedded memory and interface technologies; expectations concerning the Company's execution and results; product development; achievement of IC design wins; timing of shipments of the Company's ICs; predictions concerning the growth of the Company's business and future markets; and business prospects, strategies, objectives, expectations, or beliefs.

Forward-looking statements made during this call are subject to the risks and uncertainties that could cause actual results to differ materially from those projected.

Additional information concerning factors that could cause actual results to differ materially from any forward-looking statements made during this call are contained in the Company's most recent report on Form 10-K, filed with the Securities and Exchange Commission, in particular in the section titled Risk Factors, and in other reports that the Company files from time to time with the Securities and Exchange Commission.

MoSys undertakes no obligation to publicly update any forward-looking statement for any reason except as required by law even if new information becomes available or other events occur in the future.

Thank you for your attention. I will now turn the call over to Len Perham, CEO of MoSys. Please go ahead, sir.

Thank you, Bev. Good morning, everyone, and thank you for joining us this morning.

On today's call, I will review our first-quarter progress from a business, operational, and product perspective. Following my remarks, Jim will discuss our financial results in more detail. And then we'll open the call for questions.

I'm pleased to say that just over a quarter into the year, the Company's performance has been in line with our 2016 goals -- that is, substantially increase revenue, expand our product portfolio, add to our design-win base, reduce costs, and increase our customer count. More specifically, let's discuss our first-quarter progress within the context of our 2016 goals.

To that end, we shipped meaningful volumes of production units and a fair quantity of prototypes into new opportunities (technical difficulty) customers in this quarter. As a reminder, we have a significant base of BE2 design wins, most of which have yet to commence production.

I may mention this a couple of times. Most of the design wins to date have all been either 2012 that [flipped] into 2013, 2013. We haven't really seen any of the 2014 or 2015 design wins start ramping yet. Further, a portion of BE2 shipments this quarter represent (technical difficulty) releases from a large annual order placed on MoSys by our tier-one customer earlier this year.

We currently have a substantial number of design years for this tier-one network equipment supplier, and we continue to leverage this key relationship for additional design-win opportunities, both for the Bandwidth Engine and LineSpeed product families.

During the quarter, we also saw an increase in orders for Bandwidth Engine 1, as well as early production and prototyping orders for our increasingly large [LineSpeed] IC families.

With an increasing number of prior design wins ramping into production, we are entering the second quarter with a growing book of orders, which already exceeds our first-quarter product revenue. Though we continue to believe our full-year revenue will be back-end loaded, these early indications of a growing revenue base are quite encouraging.

We still expect 2016 revenue to be primarily driven by Bandwidth Engine products, with a lesser incremental contribution from the LineSpeed family. The number of customers and projects we are shipping to is increasing, but again, it's important to note that to date, only our initial early design wins are moving up the ramp into production.

All that said, we are pleased to see that at this early stage, the current production ramps are roughly aligning with our in-house revenue models.

With regard to new design wins for the quarter, we secured additional wins for both the Bandwidth Engine and LineSpeed families, including three with new customers. Additionally, we entered the second quarter of fiscal year 2016 with a strong pipeline of potential new design wins, spread across multiple customers.

In particular, interest in our BE2 product family continues to grow, with new market segments finding new applications, as the network infrastructure and the data center move to higher data rates and start to experience the same memory bandwidth and access challenges that come with aggregating 40- to 100-gig data flows.

These pending wins are at various stages of development but give us an increasing level of confidence for achieving our full-year goal of growing design wins in 2016 substantially over design wins won in 2015.

A little bit about industry trends. While we expect design-win momentum to continue to increase throughout the year, we have seen design decisions take longer than we've seen in the past few quarters. This is possibly due to a general sluggishness in some areas of the data [com] markets, or it could be a slowdown driven by our customer base, now starting to embrace the next generation of larger, more powerful FPGAs currently being released by both Altera and Xilinx.

The interest is [fully] there to move forward with next-generation designs. However, it seems the transitions into developments, which is where we declare the design wins, are moving more slowly, a bit more premeditated than we experienced last year.

The industry trend toward higher speeds and increased processing per packet throughout the network strongly supports the need for our Bandwidth Engine solutions in next-generation networking systems.

Applications in the 100-gig to 200-gig range would be [to have] significant value are showing up in access, security, and data-center markets; and serial memory at (technical difficulty) speeds is gaining broader acceptance and more widespread use.

Furthermore, the adoption of 25-gig Ethernet will drive rates up and create new opportunities for our LineSpeed family of [physical] interface devices. We are seeing more applications moving into that space, fueling additional opportunities for BE2, as well as new synergistic opportunities for both BE3 and BE2, serving different functions on the same line card. These are but a few of the industry trends that we believe will drive increasing design-win opportunities in the direction of MoSys going forward.

Turning now to a few product-specific updates. BE1 and BE2 are both in production, shipping to various customers. However, BE2 does represent the lion's share of these shipments.

Looking out over the balance of the year, we note that the design wins most likely to ramp imminently are BE2, not BE1, which will benefit our gross margin, given the high [area speed] and lower cost to produce BE2s.

Customer time to integrate BE2 is lessening, but it's still dependent on the customer's overall system architecture. We are also experiencing substantial customer interest in our Bandwidth Engine 3 family of products. We are having ongoing product reviews and applications leading to multiple customers who are evaluating the features and capabilities of these new, more advanced, highly intelligent networking memories.

BE3 dramatically increases the BE family's performance, with twice the serial throughput and on-chip memory capacity, and on-board intelligence, enabling increased statistics and metering capability to be brought on board.

To date, we have released two unique BE3 products, which are available for immediate sampling. We are currently in the later stages of characterizing BE3 and working on fine-tuning various other configurations, while moving towards full carrier-grade qualification.

Based on past experience, we expect that BE3 will achieve a full carrier-grade quality [of strength] and reliability certification sooner than did either BE1 or BE2. However, it is good to remember that the high-temp operating life test, for example, individually take 1,000 hours; and there's no shortening that requirement.

Our longstanding partnerships with both Altera and Xilinx have been very gratifying. The [bring-up] time to integrate these three FPGAs with either of these excellent programmable device manufacturers will be shorter than for previous versions.

This is because the ecosystem developed for the serial memory and the Bandwidth Engine in particular is significant and proving extensible between the generations based on a new-proven expertise, with low latency, high signal integrity, and high-speed [series] I/Os.

We are expecting our FPGA partners to roll out their new advanced, next-generation FPGAs in the second half of 2016, leveraging new capabilities and with [SerDes] speeds running up to 28 gigabits per second.

This should enable us to help our [shared] customer find an optimum next-generation solution faster than before, thus positioning us to shorten the time to achieve a design win for BE3, and by doing this, to shorten the customer's time to first product shipments.

We also continue to make good progress with our partner Mellanox, as they bring to market the NPS 400 processor, for which we developed our optimally configured networking memory device, the BE3 Z30.

We are excited to be partnered with Mellanox, who completed its acquisition of EZchip in the first quarter. Our relationship with the expanded Mellanox EZchip team is proceeding very smoothly. The working relationships are excellent.

We are actively engaged with multiple prospective customers for the BE3 Z30 and expect expanded opportunities in giving Mellanox this strong data-center market presence, combined with the existing strength of its acquired network-processing products and [carrier] applications.

During the first quarter, we also launched our programmable search engine platform, which utilizes much of the same interface logic and ecosystem as the Bandwidth Engine, but provides for programmability of custom capabilities to customers who might desire to implement their own functions, such as an algorithmic TCAM, specialized search, and/or numerous other applications.

Our PSE products provide MoSys with a base for new products that will precisely target new and different applications, without requiring a costly (inaudible). While both BE3 and PSE utilize a common interface infrastructure, the architecture and capabilities available to the customer with a PSE are quite different than with BE3, where functions are more defined and more standardized, in order to address specific applications.

Our PSE products behave like acceleration engines, designed and optimized to accelerate customizable subroutines, including packet processing or security functions offloaded from a network processor, FPGA, or application-specific standard products.

We are in initial discussions with customers looking to use the PSE's programmability to perform a number of proprietary functions. This product has quickly generated a fair bit of excitement and interest among both our existing customers and our larger design-win partners group.

Turning to our LineSpeed family of products, we've substantially advanced our efforts on these products and secured multiple new design wins in Q1.

Additionally, we recently introduced our LineSpeed Flex 100G full-duplex re-timer, the MSH222, supporting forward error correction and a total of 8 independent lanes, each supporting rates over 28 gigabits per second, with options to support different line cards and backplane reach requirements.

We already have multiple customer engagements for this new product, and we are optimistic about converting these into design wins over the next quarter or two. We do expect additional traction for these devices during the year, with the potential for (inaudible) contribution as we move closer to or into 2017.

It is also particularly gratifying that we have begun to see new synergistic opportunities for the use of both our LineSpeed re-timers and Gearbox devices on board with Bandwidth Engine 2 and 3 ICs, all sitting on the same line card.

Standalone or in combination with another product family, I believe our product roadmap is producing a very cohesive catalog of [leading edge] devices from which we will benefit, both in the near and in the longer term.

Before I pass this call on to Jim, I want to speak to another important 2016 goal -- improving our financial performance and reducing our cash burn.

Jim will go into more detail, I'm sure, but I'm pleased to note that we began to benefit from our restructuring and other [cost-down] initiatives in the first quarter. We are now a leaner and more cost-efficient organization, and we expect to see significantly lower annual operating expenses as we go forward.

Despite these reductions, we have ensured that we have the right resources on board to both meet customer-support requirements as they bring our products to market embedded in their products, and support new design-win and product-development activities. We are also benefitting from back-end manufacturing improvements, aimed at lowering our costs and further increasing our gross margins.

Reducing our cash burn while simultaneously ramping revenue remains an ongoing first-priority goal. We continue to pursue external collaborative programs and/or joint projects with multiple partners to fund product development and to provide some cost [relief] until our revenue ramp gains a little more altitude.

In summary, 2016 will be a pivotal year for MoSys, as we strive to achieve our goals to increase revenue, expand our product portfolio, add additional design wins, reduce costs, and broaden our customer base.

I expect our early Bandwidth Engine 2 design wins and increasing backlog to drive increasing revenue growth in the second half of 2016 and into 2017. Furthermore, our expanded design-win pipeline and new products will serve as the engine, driving continued revenue growth over the longer term.

Although still early in the year, I am encouraged by our progress to date and believe we remain on plan to achieve our goals, while working to gain additional momentum throughout the remainder of this year.

This concludes my prepared remarks. I'm now going to turn the call over to Jim for a review of our financial results. And following his remarks, we will take some questions, and I will close the call with a few comments. Thank you very much for your time and attention. Jim?

Thank you, Len, and good morning, everyone.

During the course of my comments, I will make several references to non-GAAP numbers. Unless otherwise indicated, each reference (technical difficulty) stock-based compensation expense, amortization of reported intangible assets, and restructuring charges.

These non-GAAP financial measures and reconciliation of the differences between them and comparable GAAP measures are presented in our press release and related current report on Form 8-K, which was filed with the Securities and Exchange Commission today and can be found at the investor-relations section of our website.

Turning now to our first-quarter 2016 results. Total revenue was $1.5 million, compared with $1.6 million in the fourth quarter of 2015 and $0.8 million in the first quarter of 2015.

Product revenue from the sale of our integrated circuits was $1.1 million in the first quarter, consistent with the fourth quarter of 2015, and compared with $0.2 million in the prior-year period.

IC sales increased to 77% of total revenue in the first quarter, compared with 70% in the previous quarter. IC sales are primarily driven by shipments of early -- excuse me, production and prototype build orders of our Bandwidth Engine ICs, as well as early production, prototype, and sample orders of LineSpeed units to new platforms.

Royalty and other revenue for the first quarter of 2016 was $0.3 million, compared with $0.5 million in the previous quarter and $0.6 million in the year-ago period. Royalty and other revenue is primarily comprised of royalties received from semiconductor customers whose products include our IP.

GAAP gross margin was 41% in the quarter, compared with 45% in the previous quarter and 69% in the year-ago quarter. The sequential and year-over-year decrease in gross margins was primarily due to product mix, as product shipments increased as a percentage of revenue.

In terms of our operating expenses for the first quarter, total operating expenses on a GAAP basis were $7.4 million, compared with $7.2 million in the previous quarter and $8.5 million in the year-ago period.

First-quarter 2016 total operating expenses included $0.7 million of one-time charges, related to our restructuring and cost-reduction initiatives, as well as $0.6 million in stock-based compensation and amortization of intangible-assets expenses. This compares with stock-based compensation and amortization expenses of $0.8 million in the fourth quarter of 2015 and $1.4 million in the first quarter of 2015.

Research and development expenses in the first quarter decreased sequentially and year-over-year to $5.2 million, compared with $5.6 million in the fourth quarter of 2015 and $6.9 million in the year-ago period. The sequential and year-over-year decreases in R&D expenses reflect lower personnel costs and stock-based compensation expenses.

Selling, general, and administrative expenses for the first quarter were $1.5 million, compared to $1.6 million in the previous quarter and in the year-ago period.

On a non-GAAP basis, total operating expenses for the first quarter of 2016 were $6.1 million, which includes the restructuring charges, amortization of intangible assets, and stock-based compensation expense. This compared with $6.4 million in the previous quarter and $7 million in the year-ago period.

As previously mentioned, we initiated a restructuring plan in January to reduce expenses and cash burn, as well as realign resources. We have substantially completed these restructuring activities, and we expect to realize approximately $3.3 million of annualized savings. We have substantially completed the development of our new products and will be focused on ramping revenue from existing products and winning designs for both our existing and new products.

We are targeting non-GAAP expenses to approximately $5.2 million a quarter, excluding any tape-out shuttle expenses or significant back-end expenses.

On a GAAP basis, net loss for the first quarter of 2016 was $6.9 million, or $0.10 per share, compared with a net loss of $6.5 million, or $0.10 per share in the prior quarter, and a net loss of $8 million, or $0.15 per share, for the first quarter of 2015.

On a non-GAAP basis, the net loss for the first quarter of 2016 was $5.6 million, or $0.08 per share, which excluded the restructuring charges, stock-based compensation, and amortization expenses totaling $1.3 million. This compared with a non-GAAP net loss of $5.7 million, or $0.09 per share, in the previous quarter, and a loss of $6.5 million, or $0.12 per share, in the year-ago period.

Net loss per share for the first quarter of 2016 on a GAAP and non-GAAP basis was computed using approximately 65.7 million weighted average shares outstanding.

Now turning to the balance sheet. On March 31, 2016, our cash and investment balance was $20.9 million, compared with $20.2 million at December 31, 2015.

During the first quarter, we completed a private placement for $8 million in 10% senior secured convertible notes, which are convertible at $0.90 per share and due on August 15, [2018]. The transaction was completed with existing shareholders, with no [banking fees], yielding net-cash proceeds of $7.9 million.

The terms of the notes include minimal covenants, no [warrants], a cap on anti-dilution reset at $0.85 per share, and payment of interest in kind at the company's option. We encourage investors who have not done so already to review the specific terms of the notes (technical difficulty) filed on March 15.

Cash burn in the first quarter was approximately $7.2 million. The cash burn for the first quarter of 2016 included $0.6 million of restructuring-related expenditures and $0.4 million in fixed-asset expenditures. We expect remaining cash expenditures related to our restructuring activities to be insignificant for the remainder of the year.

In the fourth quarter of 2015 and first quarter of 2015, we made significant capital expenditures to buy new hardware to support the expected increases in customer production and to provide manufacturing cost reductions. We do not expect to incur other meaningful capital expenditures for the remainder of 2016.

Our restructuring and other cost-reduction activities have positioned us to meaningfully reduce our expenses and cash burn over the coming quarters, especially combined with our expected growth in product revenues and improved gross margins.

We continue to evaluate our cash position as we evaluate our progress against our financial model but have no immediate plans to raise capital. As of March 31, 2016, our worldwide head count was 69 employees, compared with 104 as of December 31, 2015, with the majority of our employees in [applications], engineering and operations.

This concludes my prepared remarks. At this time, we would like to open the call for a Q&A session. Operator?

(Operator Instructions) Gary Mobley, Benchmark.

Morning guys (technical difficulty). Jim, I want to start out by trying to pin you down on your previous comments, with respect to the fiscal-year 2016 revenue outlook. I don't remember the exact amount, but it was -- the expectation was somewhere around $9 million or $10 million of revenue for the year. Is that what you're currently looking at and feel comfortable with today?

On the last call, I think both Len and I -- we specifically focused on IC product revenue and said that we expected it in the range -- to be in the range of $[7.9] million for the year. It's not something I'm planning on reporting on quarter over quarter, again given our lack of visibility, and really don't want to provide guidance, given it's still early in the ramp process. But I would say at this point, the comments both Len and I made were on track with our model. And we thought the IC revenue potential for the year would be back-end loaded. So we're not backing off that number today.

Okay. Obviously, the royalty revenue was down significantly on a sequential basis. And I know it's not the long-term future of the Company, but it does speak to the cash burn for the balance of the year. And still regarding that particular revenue metric, how do you feel about some improvement off of the Q1 $331,000, or should we view that low level in Q1 as the new sustainable rate?

You know, it's something we don't get forecasts on. It's primarily driven by two royalty payers with a host of other smaller ones after that. And as I've indicated previously, not a lot of new ones to turn on, although there are a couple out there we still chase with letters and all that, that may yield something.

Conservatively, given, as you point out, every dollar of that royalty revenue drops my bottom line, I tend to adjust my forecast to match the last quarter. Certainly with the beginning of the year and a new plan, I've brought it down to that kind of level, just because I don't have visibility, and it's the more conservative tack.

But I certainly hope it can bounce back. We haven't seen all the reports yet for Q2, so still in the dark there. They're still rolling in, as we're early in May.

Okay, great. And your product gross margin, I understand, is not optimal. It's, for the past couple quarters, been about 20%, and I know you've been making -- attempting to reconfigure your cost structure, and hence that's the reason for the CapEx in Q4-Q1. And so with that behind you, how do you view the progression and ramp-up of the product gross margin for the balance of the year?

I'll make an operational comment, and then Jim may or may not want to add to it.

It seems to me that we have a good understanding now of how to get the (technical difficulty) process flow of costs down -- and that would allow us a gross margin, I'm going to say, somewhere between 44% and 49%, something like that, on, I'll say, Bandwidth Engine 1 and 2 combined, Bandwidth Engine 2 probably the stronger of the two.

We can get another -- maybe as many as 10 points by going to a higher-volume [mass set] and that mass set [through expenses]. Jim and I are keeping an eye on the cash right now.

But when we get that mass set in, I would think we're going to be somewhere between, maybe 57%, 58%, and 65% gross margin. And we're [sole source] into the market with our products. They're very, very high performance, and they're very important to the solution the customer is bringing to market. So we wouldn't expect very much in the way of price erosion.

So I think, just to summarize your question, Gary, I think we should expect to get in the middle, maybe the higher-middle 40s, and then with a mass set and the few things we'll have to do, we should see ourselves get somewhere between -- I don't know, maybe 56%, 57%, and 65%. And I'm of the opinion that the product needs to be better than 60% gross margin product, based on the value in the system, and we haven't [moved off that].

And we have a good understanding of how to get (inaudible). Jim mentioned, we bought some hardware to increase our ability to handle increasing volumes of products being demanded by our customer base, and so forth. And we're in the process of getting the last things done to take as much cost out of the line as we can. And then we'll be standing by for that mass set.

Okay.

I would just add to that, Len, that we are -- under GAAP, we are still -- even when you reduce standard costs, you bleed in the change over time, over the period you ship. So we're still using the previous standard costs, just kind of drop it in a quarter so the cost structure while we're shipping is still at the higher standard cost. I do think we'll see notable improvements in Q2.

We did end up [canceling] some units in Q1 that (inaudible). We just concluded that it was too expensive to fix, etc. But I do expect to see improvements in Q2.

Okay. Last question for me -- there's been some consolidation in the communication equipment group, specifically with your customer base, and if I'm not mistaken, maybe one of your tier-one customers has recently consolidated with another. And I'm wondering if that's caused any interruption in design-win activity and/or your expected revenue ramp with that tier one.

No, I think that the divisions of our tier-one customer that we do business with, if you will, are in a business that, as far as I can see, was not being prosecuted by the emerging partner, if you will. And better yet, the combination of the two companies exposes this group to a much larger total market than he was looking into before. If you will, the customer list is going to be considerably wider.

So we haven't seen any negative effects from that merger. And it's my opinion that as the smoke clears, the group that we do business with, which I believe was the most profitable group inside of this $7-billion, $8-billion, $9-billion, $10-billion company -- I think their prospects will only get better. So it's not an issue that I can see in any respect at all, Gary.

All right. Great. Thanks, guys.

Thank you, Gary.

Krishna Shankar, ROTH Capital.

Yes. For the March quarter, you had revenues -- your IC revenues were comprised of both BE1 and BE2, I take it. And can you give us some sense for how much of the revenues came from your top-tier customer versus a broadening of demand from other design wins that you have, in terms of revenues?

So first off, on a Q1 (technical difficulty) I'm going to say that (technical difficulty) I haven't looked to see what we've done for the first quarter or what's on the backlog in this quarter. For the year, we would expect that tier-one customer to be something in the area of 30% to 33% or 34% of our revenue.

We have a large annual order on our books that got placed some time ago. So -- and I don't pay attention to [releases against that]. Typically, they end up ordering more than they [put on you] so that you end up with something greater than that. But they placed quite a large order on us here a while back.

So I would say that -- and if we look down into 2017, I'm going to make a guess that that 30% to 33% or 34% might drop down into the high 27%, 28%, and 29%, or maybe even bit more, I suppose; but that's very speculative. However, it's something in that range. That's what the percentage of the business going to our tier-one guy is, Krishna.

Okay. And then, are you shipping production BE2 now both to high-speed telecom data center and security applications? Or can you give us some sense for the end-market demand for BE2?

Just a brief comment on that. The early design wins are mostly into the core and edge [routing], and I think that -- you know, the first time we're going to see security appliance guys start ramping up might be the last few months of this year.

And if we looked at our -- you know, we've talked about, here and there at conferences, about the fact that -- from the time when you order, you [ought to just] move out 24 months and then allow for some increasing prototyping. And maybe over the 24th month to the 30th, you ought to see your customers start ramping up.

If we're on schedule, our security guys will probably start that kind of a ramp near the end of this year. Our data-center guys are about on the same trajectory, if you will.

So if we backed up, that would say that those orders were won in the second half of, say, 2014, or something like that. So it's gone through 2015, 2016. We should start to see that ramping later this year, Krishna.

So this year, just to sum it up, is mostly going to be in the network infrastructure, itself -- core routing and maybe -- I'm sorry, edge routing, perhaps a little bit of core routing.

Okay. And then do you anticipate any revenue contribution from LineSpeed this year, and if so, which particular product line seems to be getting the most traction within LineSpeed?

We're doing sampling across a fairly wide area. There's a lot of reference [boards out], if you will. I would say that -- I believe that Jim and I have been reasonably consistent that we would expect some starting to ramp in LineSpeed again in the last four or five months of the year.

I would say we've had reference boards out for the longest time on products that go into optimal modules, where the reach isn't so dramatic but the need to have very, very low power is incredibly important. I think beyond that, we're looking at some Gearboxes that had the unique features on the I/O, the forward error correction or multi-link Gearboxing, and so on and so forth.

But probably if it starts to ramp, I wouldn't be surprised if the beginning of it is into the optimal module. But it could be one of our Gearboxes -- re-timer with some unique features that aren't widely available from other suppliers, such as forward error correction or multi-linking.

Okay. And then, Jim, has your OpEx -- given all the restructuring [you've done], is your OpEx kind of stable now? For Q2, should we assume similar levels? And can you give us some sense for the cash burn going forward in Q2?

[Certainly] as I indicated on the call, I reiterated that our non-GAAP OpEx would be [just over that] $5-million range on a quarterly basis, subject, as I said, to any shuttles, tape-outs, major back-end -- I've included some back-end, but we've already spent a fair amount. Because obviously the back-end costs for the products that are in production are CapEx. Back-end costs for newer products that haven't been released to production hit R&D, and we already have the expenditures we need right now for Bandwidth Engine 3. And we're being judicious on LineSpeed and investing as we see more activity, etc.

So going forward, I think we've positioned ourselves with the restructuring to bring down -- we have brought down costs very meaningfully. Part of it is a function of just how much revenue can we bring in.

As Gary pointed out earlier, although it's still small, I don't like losing any royalties, since it goes right to my bottom line. Right now, we're just focused on measuring our model against revenue coming in. And as I think Len and I -- Len indicated particularly, so far we're on track.

We didn't -- it's a function of when these wins turn on. And assuming it takes kind of in that range of 24 months, you have to kind of look back and wait for it to hit and have some ramp period. So we're certainly going to keep monitoring our progress, etc. But as I believe I indicated, I expect [the burn] to come down meaningfully over the next few quarters.

Okay. Thank you.

I'm showing no further questions at this time. I would now like to turn the call back over to Len Perham for closing remarks. Please go ahead, sir.

So in closing, I would just like to make one comment. Earlier, Gary asked an interesting question. He asked what the ramification to us was with respect to an M&A action taken by one of our -- who he thinks one of our good customers might be. And I commented on that.

But I would go further and say, the part of the question that he didn't ask was -- some of these M&A activities have reduced the number of suppliers of niche-type products that are very, very valuable to the customer base we all serve.

By that, I mean it could be that in some of these M&A activities, where there was once two suppliers, there is now one. And in some of these M&A activities, it might be that the two companies combine and they're pretty huge, and their strategic plan going forward for product families and where they want to prosecute their business is such that they don't have interest in niche products anymore.

And we are seeing some real interest in that area. My background [through] my life was building these niche-oriented products that are not widely available but provide great value in the system. I think IDT might have been maybe 75% sole-source when I was running the company. And clearly, NetLogic had many, many products that [the only] guy in the world delivered. So it's an area that I'm familiar with.

And I would say, the most interesting thing we're seeing, perhaps, as a result of all this M&A activity is the consolidation of small niche players into one very big player, or acquired by a big player, who may leave the customer with one supplier, or may not even have one supplier because it may be [obvious the strategic intent] of this much larger enterprise is to not serve that business at all.

And uniquely architected networking memory is very, very important to the system going forward, and it's going to be more important in the future than it is today. It's critical. And the Bandwidth Engines are very, very innovative, and they support a very high performance. And it's an area of opportunity for us.

So having said that, that's the only issue I thought I left unspoken to this morning. I want to thank you all for dialing in to listen to us. And we're hoping to let you know about even better results next quarter. I want to thank you for your time, and that's all we have for today. Thank you very much, and have a great day.

Ladies and gentlemen, this concludes today's conference. Thank you for your participation, and have a wonderful day. You may all disconnect.