Peraso Inc (PRSO) 2012 Q3 法說會逐字稿

Good morning and welcome to the MoSys third quarter 2012 results conference call. At this time all participants are in a listen-only mode. At the conclusion of today's call, instructions will be given for the question and answer session. (Operator Instructions). As a reminder, this conference call is being recorded today Friday, October 19, 2012. I would now like to turn the call to Beverly Twing of Shelton Group investor relations. Beverly, please go ahead.

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

The third quarter 2012 financial results press release was distributed earlier today and is available on MoSys website at www.MoSys.com.

Before we begin today's discussion, I would like to remind everyone that 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 27-A of the Securities Act of 1933 and Section 21-E of the Securities Exchange Act of 1934, which include but are not limited to benefits and performance expected from use of the Company's embedded memory and interface technologies and ICs, expectations concerning the Company's execution and results, expected benefits of the bandwidth engine ICs, product development, achievement of design wins, and timing of shipments of bandwidth engine 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 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 Annual Reports on Form 10-K and Form 10-Q filed with the Securities and Exchange Commission, in particular in the section titled Risk Factors, and in other reports the Company files from time to time with the Securities and Exchange Commission. MoSys undertakes no obligation to publicly update any forward-looking statements 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, Chief Executive Officer of MoSys. Len, please go ahead.

Thank you, Bev; good morning everyone. Thank you for joining us. I'll start today's call with an update on our Bandwidth Engine IC family and then discuss our progress in design wins as well as our continued development efforts for our next-generation products. Jim will then review our IT business and third-quarter financial results and prior to opening the call for a few questions.

Let me start by saying that the level of sales activity reached an all-time high in the third quarter. Additionally we are seeing increased interest from customers and other potential partners interested in exploring ways that we might improve chip to chip important performance and therefore increase overall system performance for our shared customers.

Let's talk a little bit about design wins. Looking back, we announced design wins on multiple platforms and involving several customers in the first quarter of 2012. These customers are now moving forward with board qualification and final system design activities at a reasonably at a reasonably intense pace. Because the boards have multiple bandwidth engines on boards, as many as 11 units per board in some cases, we are deeply involved in supporting the customers' board qualification and system integration efforts.

When the board qualification and system integration efforts are completed, we would expect the customer to move onto the next step. That is prototype system builds. Systems you would put into his channels to see how successful his system product will be.

Finally, we continue to anticipate the systems where we won our first designs should get released to production in the second half of 2013 going into 2014.

Supporting these boards bring-up and system integration efforts has drawn us much closer to the challenges our customers faces in bringing out a new system. It is an incredibly important process for us to participate in.

First it shows us other tools and support collateral we should be developing that will aid in shortening this bring-up period for the customer. Secondly it provides an opportunity for us to collaborate with other chip vendors supplying products to our shared customer, thus enabling us to share ideas on improving chip to chip performance and future systems. And finally it leads us to new features we can include in our future products that would further endear us and lock us into our customer relationship.

Let me give you an update on our design wins in progress. Sales activity has been gaining momentum throughout the quarter, and our progress toward further design wins continues to be very encouraging. Previous design wins allow us increased credibility and raised curiosity and interest of new potential customers.

As I mentioned on last year's call, I'm sorry; excuse me. As I mentioned on last quarter's call, the potential for further design wins before year-end continues to be a strong possibility. We are working diligently on closing these business opportunities and look forward to further updating you when appropriate.

For this morning I would tell you we have had no unfortunate setbacks. It's just the time that the normal process takes. I believe we're making good progress, but I must remind you in this market not only is the sales cycle long, but additionally the time-to-market for the customers networking system is long, too.

It's a long runway, but the product is a very long cycle time, lifecycle if you will, once you win the socket and [he's] in production. Last quarter we reported having as many as 10 design wins somewhere in progress. That number holds true for this quarter as well.

I think we're actively communicating with a wider array of potential customers than ever before and the activity level in our sales funnel is particularly gratifying. Lots of things going on, lots of push to make -- get new systems onto the table and into the tunnel to be developed and released to serve the expanding needs of the network.

We continue to leverage our current distribution channel and I'm pleased to see momentum building in our three primary sales regions -- Japan, China and the United States. Our relationships with Macnica in Japan, RTI in China and with Macnica Americas as well as local sales representatives on the East and West Coast in the United States, along with our internal efforts, are providing us with the necessary level of intensity in all of these various far-flung market sectors.

Specifically, we're making excellent progress in China and we're cautiously optimistic we might win a design there by the end of this year or early in the first half of 2013. We are seeing strong design in activities on multiple OEM platforms and continue to be encouraged by our progress in the China market.

We also have multiple design wins and design wins in progress in Japan, a country well known for developing a leading edge Internet and leading edge network system technology.

In North America we have identified new Bandwidth Engine opportunities at US tier 1 OEMs and we continue to expand level of discussion we're having with these various industry leaders.

With the increasing activity and interest surrounding BE1, it's only natural that BE2 and our advanced product roadmap would stir up a lot of interest as well. Regards Bandwidth Engine 2, we continue to make good progress in the development of Bandwidth Engine 2, the next product on our roadmap.

Logic verification has been done for some time and the effort of these last few weeks have been primarily focused on timing closure. BE2 represents a significant increase in I/O speed, coupled with a substantially enhanced feature set while the same time representing a substantial cost reduction over BE1. Additionally, a greater than 4 gigabyte access per second is clearly providing its user with more access per second than in any other IC on the market today.

Achieving this requires that we drive the internal circuitry very hard, thus the extra diligence and the extra time being spent on timing closure. It's coming together quite well. The fact that we're taking a little longer isn't going to set us back with any customers, and it will save and bring up time.

And additionally what we're learning now will bode very well for us on BE2, 2.5 or 3. Future products will benefit from the time taken to understand exactly how to be very precise on timing closure. All of this said, we have recently announced the first two members of the BE2 family.

The first, the MSR620, is optimized for buffering applications up to 384 gigabits per second and includes both write and read, and write and broadcast capability. It also accelerates larger packet sizes up to 576 bytes and has an unprecedented 80% overall bandwidth efficiency which is beyond the capability of currently available standard memory subsystems and alternative serial interface solutions. In other words, we're the benchmark for others to try to follow.

Our second BE2 -- the second member of the BE2 family now released, the MSR720 access device, is optimized for high access rate applications and [packer headed] processing, and is well suited to the requirements of state memory and queuing applications. It also contains memory coherency, which allows back-to-back read and write to the same memory bank, allowing for higher efficiency for small tables.

These two new BE2 devices, with their advanced capabilities and robust feature sets, will allow us to compete for additional sockets on next-generation line cards, thus expanding the size of our market opportunity, thus expanding our opportunity for ownership on a given line card or network interface card. We believe that these first two BE2 devices and others on the road map will allow us to become a leading enabler of the advanced 100 gigabyte and higher performance networking systems required to support the growing Internet infrastructure requirements.

While the industry is currently primarily focused on developing design for networks up to 100 gigabytes, it is also beginning to look further out at 200 gigabyte and 400 gigabyte systems as the demand for rich content processing continues to increase, fueled in part by the exponential growth of the connected devices and mobile applications, either connected to or riding on the Internet.

Over the past several months we've been collaborating with increased numbers of the leading network and communications equipment companies to identify the significant design challenges and expected requirements for these 200 gigabyte and 400 gigabyte systems. This collaboration has allowed us to become involved in the early stages of the design process, and we are leveraging this knowledge to offer MoSys design solutions that will meet the speed, performance and low power requirements of these new future generations of networking systems.

Further, these discussions have provided us with the opportunity to demonstrate how our technology and its capabilities can be critical to the success of these next-generation networks because many of today's solutions are simply incapable of meeting this demanding level of performance.

BE2 was ideally suited to solving many of the systems problems associated with 100, 200, 400 gigabit levels of performance.

Our strong product roadmap provides for additional BE2 derivatives and additional generations of Bandwidth Engine ICs that will enable us to capitalize on the future transition of the network to 200 gigabit and 400 gigabit networking systems.

As we introduce our BE2 products in the coming months for also looking at futuristic line cards and network interface cards to identify other products and development opportunities that would leverage our extensive analog design expertise.

We are exploring these other board level elements that we can consume that will produce higher performance with a reduced chip count or provide higher performance with the same chip count at a lower cost per unit of performance to our customers.

In conclusion, it has been a busy quarter with intense activity and gratifying progress on nearly every front. We've experienced increased activity with customers and prospective customers, advanced our design wins toward production moved our second-generation Bandwidth Engine family of IC towards to our [safe house].

Our sales efforts are gaining momentum with current market opportunities, and we believe we are well positioned as the market performance requirements expand with opportunities beyond 100 gigabyte, including 400 gigabyte and beyond.

Our customers and partners are beginning to see the full capabilities of our ICs and our technology. As we continue to leverage our expertise and solutions to solve the bandwidth and access rate challenges confronted by network and communication systems, we are taking measurable steps toward generating meaningful Bandwidth Engine revenue and further supporting our objective of becoming a fabless semiconductor Company.

In short, we had a pretty good quarter and we're looking forward to another one now.

With that, I'm going to turn the call over to Jim to discuss our IP business and financial results, and then we will open the call for questions. Thank you very, very much for your time and attention.

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 will be to an amount that excludes stock-based compensation expense and intangible asset amortization. These non-GAAP financial measures and a 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.

Prior to discussing our financial results, I would like to first address our IP business. We continue to record licensing and royalty revenue from ongoing Certes and 1T-SRAM IP projects. However, as mentioned on previous calls, we are not actively pursuing additional licensing agreements as we've completed IP asset sales over the past few quarters.

This aligns with our strategy to deemphasize and monetize our IP in order to achieve our primary business objectives of becoming a fabless semiconductor Company. That being said, our team continues to support our existing IP customers and we continue to record revenue from ongoing projects.

During the third quarter, we completed the technology transfer obligations to Synopsys, which arose from our $4 million sale of Certes IP in the first quarter of 2012. Upon completion of these obligations, we received a $1.3 million milestone payment.

All technology transfer obligations related to this transaction were completed smoothly and on schedule due to the commendable collaboration and efficiency displayed by both MoSys and Synopsys teams. We expect to receive an additional $0.6 million payment from Synopsys in the first quarter of 2013.

As we have completed the technology transfer obligations to Synopsys and our performance obligations under our existing one 1TS-RAM and Certes licensing agreements are winding down, we have fewer projects for our team to work on. Yesterday, we initiated a 10% reduction in our workforce and eliminated 10 positions in Santa Clara, primarily in engineering-related functions. We'll also be closing our small office in Taiwan.

As a result of this reduction in force, we expect to record termination related charges of approximately $0.3 million during the fourth quarter of 2012. Excluding the effects of one-time charges, we expect to save approximately $2 million in costs on an annual basis from these headcount reductions. Now let's review our third quarter results.

Total revenue was $1.3 million compared with $1.7 million for the second quarter of 2012 and $2.1 million for the third quarter of 2011. Licensing and other revenue for the second quarter was $0.2 million compared to $0.6 million for the second quarter 2012 and $0.8 million in the third quarter of 2011. License revenue was comprised of follow-on revenue from existing licensing agreements and maintenance renewals. License and other revenue for the third quarter included revenue from 7 licensees compared with 10 in the second quarter.

Royalty revenue for the third quarter was $1.1 million compared with $1.1 million for the previous quarter and $1.4 million for the third quarter of 2011. Third-quarter royalty revenue is primarily comprised of royalties from a foundry partner and a fabless semiconductor company selling into networking systems. Third-quarter 2012 royalty revenue was recognized from 14 licensees compared with 12 licensees in the second quarter of 2012.

GAAP gross margin for the third quarter was 96% compared with 90% for the prior quarter and 83% in the year ago quarter.

In terms of our operating expenses for the third quarter, research and development expenses were $7 million compared with $6.7 million in the previous quarter and $6.6 million in the third quarter of 2011. Research and development expenses increased sequentially primarily due to expenses related to continued product development costs.

Selling, general and administrative expenses were $1.7 million compared with $1.4 million in the previous quarter and $2 million in the year-ago period. Total GAAP operating expenses for the third quarter was $7.3 million and included the asset sale gain of $1.4 million which included the milestone payment from Synopsys, $0.3 million for amortization of intangible assets and $0.9 million in stock-based compensation expense.

Excluding the net gain of $1.4 million on the asset sale, total operating expenses were $8.8 million compared with $8.1 million in the previous quarter and $8.6 million in the third quarter of 2011.

On a non-GAAP basis, total operating expenses for the third quarter of 2012 were $6.2 million compared with $6.6 million in the previous quarter and $7.1 million for the third quarter of 2011. We continue to closely manage our operating costs while maintaining the necessary investment for development of our Bandwidth Engine integrated circuits.

On a GAAP basis, net loss for the third quarter was $6.1 million or $0.15 per share, compared with a net loss of $6.6 million or $0.17 per share in the prior quarter and a net loss of $6.9 million or $0.18 per share for the third quarter of 2011.

On a non-GAAP basis, net loss for the third quarter 2012 was $4.9 million or $0.12 per share, which excluded intangible asset amortization and stock-based compensation expenses totaling $1.1 million, compared with a non-GAAP net loss of $5 million or $0.13 in the previous quarter and a loss of $5.3 million or $0.14 per share in the year-ago period.

Net loss per share for the third quarter of 2012 on a GAAP a non-GAAP basis was computed using approximately 39.3 million weighted shares.

Now turning to the balance sheet, as of September 30, 2012, our cash and investments balance was $45.3 million compared with $48.9 million at June 30, 2012.

As of September 30, prior to the reductions I discussed earlier, our total headcount was 107 employees compared with 143 employees as of December 31, 2012. Over 80% of our employees are in engineering and research and development. And of our total employee count, approximately 18 are located in India compared with approximately 20 in the previous quarter.

This concludes my prepared remarks. At this time we would like to open the call for a question-and-answer session. Please clearly state your name and company affiliation prior to asking your question. Operator?

(Operator Instructions) Gary Mobley, Benchmark.

Hey guys, happy Friday.

Good morning, Gary.

Len and Jim, I realize that it's difficult to distinguish between design wins in progress and actual design wins. Len, you mentioned 10 design wins in progress. I'm curious out of those 10, how many do you think will actually make it to the prototype system built phase by the second half -- well, actually by the first half of 2013?

Well, Gary, let me just say that some of them are already -- have already received a fair number of reference voids and sample quality quantities of components. So that would sort of underline the fact that we are conservative in what we bring, the information to our investors and shareholders.

And we haven't got to this stage -- I would say that our batting average when we get to this stage is maybe 60% to 70% or something like that. We haven't had a lot of losses when we get down this far.

I think to our favor, as we move from 100 gigabyte to 200 gigabyte, to 400 gigabyte platforms I think it's becoming increasingly more obvious that there might not be any other way to accomplish some performance levels as you move up from 100 gigabyte to 200 gigabyte to 400 gigabyte without our approach. We are seeing that customers are really beginning to appreciate the power and the capabilities of the device as they look toward their future boxes.

So without trying to be super precise, we've had a setback here and there. Usually when we get a setback, the guy decides if the switchover costs are too high and he can find a way to bolt it together using his old approach. I don't think -- I want to be careful, we had a -- as I reported in the past, we lost a deal or two here and there when a customer decides to do a customer-specific chip. And usually that would've been in force before we got there. We've tried to catch up with boxes that are just too far along in the guy doesn't want to take on the extra risk of missing his time to market.

But all in all, as we move up the performance level we're getting involved earlier and earlier. And I think we should expect a fair number, maybe more than half, should get across the goal line.

Okay, you mentioned shipment of some reference board to some potential customers and the next step would be prototype system builds for these customers. When they enter their prototype system build phase, how much more incremental revenue might you see from the sale of Bandwidth Engine? I know we're talking several millions a quarter, but how many units during that phase?

Well, that's a little hard to project because some of these guys are -- depending on where the box were being targeted for is in their queue of getting things out, one might move faster than the other. And probably, you know, the only way to do that would be to have Jim talk about what he is projecting revenue to be, perhaps, for 2013. And that might be a little bit long for this call.

We're projecting thousands of units to sell and [10,000] in 2013 but not tens and hundreds of thousands, right? So I may be off. Maybe it is going to be more than 10,000. But it isn't going to be earthshaking and each guy is moving at a different pace.

Okay, I have some questions for Jim. Just to be clear, the milestone payment from Synopsys for their IP purchase, that was recognized as a gain on sale. Is that correct?

That's correct. It was -- I rounded it [in the fall]. It was actually we reported a gain of approximately $1.4 million, of which $1.25 million was a cash payment. Then it was a small non-cash gain we took because our estimate of project hours to complete the transfer obligations came in under plan. So we released the cost we had set up on the balance sheet.

So the cash burn in the quarter, had you not received this milestone payment, would have been roughly $5 million?

Yes.

Okay. So, looking into the fourth quarter, I'm assuming with additional tape out expenses for Bandwidth Engine 2, you know I don't know if that involves both of the product tape outs, but we should probably assume a much higher R&D level in the fourth quarter than a commensurate much higher cash burn rate?

That would be true. We're projecting the tape out to run somewhere in the $1 million to $1.5 million range. We actually have brought down the cost.

And I won't -- I'll botch it; I forget the term we are using, what we're doing, but we've dropped the cost down. But your assumption would be would be reasonable.

We also did have higher option exercise activity in the third quarter, which obviously contributed to reducing the cash burn. And it's hard for me to project will happen in the fourth quarter, but the option exercises were pretty high. But I think you're spot on in what you're thinking given the undertaking we have here with getting the tape out done for Bandwidth Engine 2 in the fourth quarter.

Sure, last question for me. With the Nintendo Wii you having rolled out a few months ago, I'm assuming you're no longer receiving any royalty contribution from the old Nintendo Wii platform?

I wouldn't say we're not receiving any, but it has dropped significantly. And as I think as I said on previous calls, they now report -- Nintendo, I should say, not they. Nintendo reports Wii and WiiU together in their quarterly press releases and public filings. So I've lost visibility.

But based on what I've tracked to date of how many Wiis Nintendo has shipped versus how many units we shipped to our customer, you know, I expect that to maintain at a low level. I'm also waiting to see what price-cutting Nintendo may do on the traditional Wii here with the holiday season coming and see if that spurs any growth. What I see based in the channel already that's -- I don't expect that to come back anywhere near to the levels it was.

All right, thanks guys.

Thank you Gary.

Krishna Shankar, Wells Capital.

Yes, can you talk about the revenues that you expect in 2013? I think it will mainly come from Bandwidth Engine 1. And also can you give us some sense for the design wins in Bandwidth 1 and the types of customers there versus the types of customers you're engaged for Bandwidth Engine 2?

So I don't have a system diagram sitting right in front of me, but we're basically pretty much spread through the system. We're heavily in the core and moving out from the core. In some cases we do get out on the edge. So we're finding our way into their switching and routing opportunities spread across the system. Wherever you see systems running at very, very high speeds then you're going to find a way into those boxes.

As far as Bandwidth Engine 1 and Bandwidth Engine 2 are concerned, we dramatically increased the feature set in Bandwidth Engine 2. In one case, we put a cache memory in it that allows it to do burst memory functions. And that's allowing us to target a few -- some sockets on the line card we had previously not looked at.

We have substantially increased the features, or I should say the robustness of the memory in terms of its ability to solve some QN policy and statistics requirements, where it might have taken to BE1s. If you look down the road a couple of generations, you could probably do it with one BE2.

We're running the I/O 50% faster so it's going to be quite desirable for advanced systems at 200 gigabyte and 400 gigabytes, because of getting in and out of the memory is vitally important as you jack up the performance. And so by running the I/O up around 15 gigabytes instead of 10, we're able to have memory accesses at 4 to 5 gigabyte access per second up from 2.75. So in future systems that are going to run at these new levels of performance, the BE2 above a certain point would be a much more ideal solution.

Just to make sure I touch on it, I don't fear that it's going to cause customers to stop in their tracks and not use BE1 but switch to BE2. Because of the I/O speeds we're talking about here and the sophistication of the solutions we're providing, once you design in BE1 he is going to be happy with it for a while, which brings us back to the point that in 2013 we'll be driving costs out of BE1 so that we can achieve the optimal margin for that part, as well as the margins we might get on the BE2, which is roughly a -- substantially lower in cost.

I think to -- I'll chime in now to address Krishna's question with regard to 2013 and Bandwidth Engine revenue. And I think Gary had alluded to it as well in his question. Unfortunately we're not in a position to give guidance. That is our policy.

I think as Len said earlier, we are shipping units. We continue to ship units in the hundreds to our design wins that we have actively working now. As we move into 2013 we expect that to move into thousands and tens of thousands, potentially, but difficult for us to predict. So I really can't give you a range given our practice, Krishna.

Okay, and you know, you said that as you move to these higher speeds of 200 to 400 gigabits there's really no alternative to your solutions for high-performance bandwidth. Can you talk about the competitive landscape?

And what would be the size of the market you're addressing in 2013 with the Bandwidth 1? Given the design wins that you have, what would the served size of the market you're addressing in 2013 and then potentially in 2014, you know, as you start to introduce Bandwidth Engine 2 in production?

So, you have asked me three questions. First off, the size of the market doesn't change. We've said that Bandwidth Engine 1 was designed and probably targeted into what we thought was the need for system designers to look beyond the RL DRAM for the performance levels they are currently moving toward. And you know, we provided some estimates of $300 million, $400 million, $500 million, $600 million TAMs for that business, if you look across the entire range of application of RL DRAMs.

By some of the -- so, some of the -- and then we also have told our customers and -- our shareholders and partners here that the Bandwidth Engine very effectively targets QDR memories in many cases, depending out where they're sitting on the line card. The addition of a cache memory probably allows us to attack that market much more vigorously, and I think the two big owners there are probably -- have been in the past Samsung and Cypress, and with the ISSI also doing well in that space.

And I think that market is probably $300 million or $400 million market, and certainly we think we could play a role there as well. So the TAM is very, very large compared to a small Company such as ourselves. When we bring out BE2, we're providing a substantially increased level of performance which allows us to look into opportunities for systems that are coming onto the design table after the ones that are embodying BE1. In some cases, depending on timing, they may be internally deciding whether to use BE1 or BE2.

But the key point is BE2 is going to be a performance upgrade and serve applications where performance is higher than where we had positioned BE1. Additionally, beyond performance, we substantially increased the feature set so that we're much more effective at going after maybe as many as 6, 8 or 10 needs for very fast and intelligent memory on the line card. So the feature set allows us to go for sockets that we maybe didn't consider on BE1 and try to take more board ownership.

Beyond that, with the increasing levels of performance that we can achieve on BE2, we can do that at approximately the same power. So for customers that are trying to get more performance in new systems, but hold the power level constant because dealing with power and the system is a very expensive process, we're also giving the customer the opportunity to get performance without having to worry about substantially more power being dissipated into his box.

And finally, by moving from one process node to another we're able to implement an effective cost-down strategy that would enable us -- and that should certainly allow us to pass some pricing relief over to our customer and put him onto a decent learning curve. So that's a lot of information; I hope I've answered your question (inaudible).

That was very helpful, thank you. And then, Jim, on the cash burn you said that Q4 we'll see an increase in your cash burn as you tape out and so on. So can you talk about the cash burn profile, how that might look through 2013? And are there any other opportunities for IP licensing or contract design revenues, or any other sort of things that you could come and -- (inaudible)?

Okay, sure. You know, from the position of where we sit today and looking out into 2013, our year-to-date burn for the first nine months is just under about $13 million. As I mentioned in response to Gary's question earlier, we will see a higher burn here in the fourth quarter. With obviously the Bandwidth Engine tape out, tough for me to predict option exercise money coming in and how that helps us. But regardless, put it in somewhere in the range this year let me just say closer to $15 million to $20 million, closer to $20 million.

I expect next year to be somewhere in that similar profile. As I mentioned during my script, we did have some reduction -- reduction in force, which was very difficult for the team here to do, but something we had to do in light of the scale back in IP and lack of projects for the team. We've also had attrition throughout the course of the year, and sometimes it's tough to see with kind of one-time items in there and tape outs and a lot of costs, et cetera. But we also had attrition throughout the year where we just had some positions hopefully to not replace them, et cetera.

But I think next year the burn -- I'd like to target it to be closer to $15 million -- in the $15 million range. But it's tough for me to see and predict items that could benefit.

As Len has mentioned on previous calls, and I probably have mentioned also, we are looking at further ways to monetize our IP. We have been very successful so far, I believe. And between the patent sale in the fourth quarter of last year and the smaller Certes deal, close to $40 million there, we're hoping there's some water left in the well. I know Len will have some comments here when I'm done speaking. But you know it's difficult to see.

We also continue to explore strategic opportunities which could be a source of cash, if we partner with somebody, maybe did something unique for them with a next Bandwidth Engine or other relationships, et cetera, although it's tough to say. And many times those types of transactions have milestone payments, and tough to predict when the cash will come in.

So, you know, this morning I didn't talk about the -- I think I have in the past talked about the fact that we're looking very hard for things, driving revenue and earlier adoption and getting revenue to start ramping up earlier, so that we start making the journey toward being a sensible business and making money. And we did talk about that a lot by talking about demand.

The second thing on these four items that are on my desk all the time are costs, and we've been looking very hard at costs. And as we've wound down our involvement in IP contracts that we had before we came across to becoming a fabless IC Company, we have taken the unfortunately hard step of scaling back the workforce and doing -- making changes necessary to make the Company of a sensible size and configuration to be an IC Company instead of an IP Company. And we'll continue to look at that.

But these very, very complex serial I/O memories to take a large group of people to create them. And we won't cut the muscle so that we can't have a really powerful long-term future.

The third item is IP that Jim just touched upon. And we -- and that sort of falls on my desk. I take care of that project primarily. And we have been in conversations with some folks who have -- not too much we can say today. We don't have anything on the table that approaches what we did in December of last year.

However, you know, we're looking at some opportunities right now. And it's just a little early to tell you folks that we've got it in the bag. In terms of scale, how we could pull in other solutions for the board or how we could look at just allowing ourselves to have more value to our shareholders by some clever arrangement, I look at that all the time and I don't see anything to report, or I don't have anything to even exciting to talk about there.

Beyond that, beyond the fact that we're looking at IP and there are a few things kicking around, I did mention on the call that some of our big customers in the tier 1 level have continued to push it for second sources. And we don't have a second source to report today, but we're spending a fair bit of time working on that and John Monson leads that charge. And I think reason to believe we'll provide our top-tier customers with a second sourcing here, so that they can take that off their list of things to worry about.

Anyway that's a long-winded answer to the back end of your question.

(Operator Instructions) Jeff Schreiner, Feltl and Company.

Good morning gentlemen. How are you doing? Thank you very much for taking my question. Sorry Jim, this is going to be mostly Len questions here.

One thing I'd like to understand, after watching MoSys's presentation, watching other companies' presentation at the Linley Processor Conference -- I believe it was last week or the week before -- how can Bandwidth Engine really kind of alleviate the memory bottleneck? And can Bandwidth Engine get the processor up to almost full bandwidth utilization, where today a lot of processors aren't likely using all the bandwidth they're able to pump through? How would Bandwidth Engine help kind of change the dynamic there?

So, actually, the answer to that question is right at the input edge of the Bandwidth Engine is the GigaChip interface, which we talk about all the time now and we encourage people to adopt. And you know, if we back up a minute and we think, well, before we arrived on the scene what was going on? Well, you know, out of dire need Cisco adopted -- or they led the charge I think to adopt Interlaken, and they may have even been the author and creator of the Interlaken interface.

And then either they or they help somebody create Interlaken Look-Aside. And this is a backplane interconnect that is typically used for managing very large packets of information. So if you're moving hundreds of -- I'm sorry, hundreds of bytes of information or thousands of bytes of information or tens of thousands of bytes of information, 64 bits or so of protocol security CRC doesn't make any difference. But when you're just trying to move 10 bytes or something like that around on the line card, or even below 100 bytes, the efficiency of Interlaken will drop down to maybe 35% or 38%. And that would say that if you're running a gigabit information, your efficiency has dropped down to 380 megabytes and/or at 37%. And that's a very big problem.

So we said that we were going to fix the packet size via the GigaChip interface to 80 bits. And 80 bits would always be used for CRC, security, protocol, whatever you need. And 72 bits will always be data. So we offered to clamp the bandwidth efficiency at 90% efficient.

As we talked to some of our big customers, or potential customers, let's put it that way, some of them have said we'd like to use a 512 bit packet size or we'd like to use the 144 or 288 or 64 or whatever. And we sit down and we talk with them. We say, look, where do you think the GigaChip interface [will we fix it at] 80, we can lock in bandwidth efficiency.

This particular problem comes up as you drive systems faster and faster, and without violating the GCI interface, which we would like to be a standard much like using the PCI Express model where people can make sensitive changes to it. But you have kind of a caretaker that doesn't let anybody's changes correct the underlying GCI interface, so that all things GCI will always run like all things GCI and there isn't one thing that causes it to corrupt.

Inside of that framework we've been able to get very, very close to 90% efficiency and match in our I/O to the eccentricities of our customers. And I've made the comment here and there that after Bandwidth Engine 1 and the introduction of the GCI interface, the Company, MoSys, really doesn't need to talk about bandwidth anymore. What we talk about now is just plain access to the memory, and that's where this 2.75 gigabit access is going to 4.5 gigabyte or 5 gigabyte access becomes very important.

There's two issues here in your question. One is the bandwidth efficiency, which I think we have in every case, allowing for normal system inefficient or eccentricities, if you will. We're running somewhere between 80% and 90%, at -- I'm going to say every time I've heard it discussed.

And then on accesses, because there's multicore processors out there that are running extraordinarily fast and they're governed totally by their ability to access memory, the fact that we're moving the gigabyte access to the memory on Bandwidth Engine 2 to 4.4 gigabyte to 5 gigabyte is really opening some eyes.

And I should mention, by the way, the round trip time from the I/O from the input in back to the output, we've shortened that on Bandwidth Engine 2 as well. So access speed had to go up to accommodate, or go down, if you will, less time. We've also addressed that. So those are the ways we achieve the solutions that your question implies.

Thank you for that. Actually I was incorrect, Jim, I'll pass this one on to you. Talked about the second half '13 release of a new system in the design win for Bandwidth Engine 1. Is that going to be revenue recognition in the second half? Or are we then going to see a qualification period of 9 to 12 months before this product really ramps?

Right now, as I've said last quarter and this call, we're shipping units in the hundreds. And the revenue is at a level where we're not going to break it out separately on our P&L.

But right now is the time we're putting hefty GAAP reserves, given that the last couple of quarters here is the first time where we're putting units that are going to go into actual systems, so there's a bit of an unknown until you see it on the actual boards coming back, rather than just onesie/twosies in test, but moving them in hundreds. And the guy building boards for his systems -- now is the time.

I expect by the time that the guy has completed his system qual and moving into production will be beyond that, and ideally we're not expecting any issues. So we will have flushed through that and the reserves will be down to what you see from a normal traditional fabless company as far as some kind of warranty reserve related to that.

And we'll get through that now here in early stages, is my expectation. You know it's the first time, so now we've got some hefty reserves to satisfy ourselves and satisfy our auditors.

I would chime in behind him and say that this is the way an integrated circuit supplier should behave, that in the beginning of a new product, and God knows we've got incredibly new product to bring to our customers, you want to just err on the side of caution. But we've had no issues out with our customers. So we're anticipating that will just come down like a learning curve and it will -- and the reserve will become relatively small.

Back to you, Len. Obviously many in the industry have talked about longer-term a lot of these memory solutions moving to more serial-based offerings. And certainly I think that maybe some of the SRAM products are not serial in nature at this point. But if they were to shift towards that direction, could that extend their life and extend their competitive offering, I guess, as an alternative to Bandwidth Engine or Bandwidth Engine 2 in future years? What's your thoughts on that?

My thoughts are that if some of the memory guys, the guys that do the QDR memories were to adopt the GCI interface and come out with a serial form of their memory, they would enable their customers to be more successful at bringing out his high-performance systems, and I would encourage them to do that. I think that not only that, they should take away these QDR memories and they should start thinking about how to build a little intelligence into them. So that you take a solution call the Quad data-rate, which I think trademarked when I was running IDT, and you move it to the next -- I'm going to call it performance node to support the shared customer.

Now, then you could say to me, Len, what's that going to do to you? We're not afraid of that. At the most advanced process nodes, which I would say are 40, 28 or 20 -- I'm just going to make a guess that we're probably one-third the cell size and one-third the power. And we're off serving some other applications on the board.

But, yes, in order for the customer to get to higher and higher speeds, he's going to have to either abandon the solutions he's using now totally and just move to a new one. Or those solutions should move with him as he tries to, you know, serve the needs of his customer base, i.e. the network itself.

So, yes, memory guys going serial and building intelligence into their products is not a bad thing and it's what the customers need. So that's the answer to your question.

Final one for me, you talked about possibly some new opportunities; maybe making announcements in early 2013 if they come to fruition. But could you give us a handicap, a percentage of where you think the opportunity is that to even bring maybe one additional customer on by year-end, or be able to talk about having another tier 1 type design win?

What do you mean when you say where the opportunities are? You mean what companies?

No, no, no; not trying to get into the companies, I believe in the prepared remarks you talked about the opportunity to engage with several new customers, and in particular, have that by year-end. I'm just trying to understand maybe a handicap on that, even if it's just one of the many customers you're talking with.

So basically the top five guys in the business are Cisco, Juniper, Huawei, ZTE and ALU. And I think I've got them out of order. I'm going to guess it's probably Cisco, Juniper, ALU or Cisco, Juniper, Huawei, ALU; something like that.

I made a comment this morning somewhere in the prepared remarks that we're encouraged about the progress we're making in China. There aren't a lot of guys serving the network in China. There's H3C, ZTE, Huawei. They're all big, powerful, pretty magnificent companies.

And so a guy could drive the line draw the line that, hey, the guys want to be working with one or more of those guys. We made the comment that we're gratified by the progress for making with tier 1 guys here in the United States. In the United States you probably could find Cisco, Juniper, ALUs. We've got a very, very strong presence here, so any one of those.

I'm not in the -- I used to tell people when I ran IDT that they were in the business of fortune-telling, and I wasn't forecasting. I was just telling the results, so handicapping isn't really my strong point. But if I didn't think we had a reasonable probability to do it, I wouldn't even mention it. Why would I tell you guys something that wouldn't come to pass?

So without putting a percentage on it, if I mentioned it, then I'd be pretty damn sure I thought carefully about it before I told you guys. Basically you guys are my most important allies. So I have to -- and I never want you to miss the line -- you know, the achievable or set expectations that aren't real.

On this call there's been a number of people asking about how fast will the revenue rise. But we -- again in my prepared remarks I caution you today, it takes a while to wind these big box designs, and then you have to be patient and work very hard with the guy to help him get his big box out. But once it's out there, it has a very, very long lifecycle and your product has a very, very long lifecycle as well.

And the biggest challenge the Company has now is not the acceptance of its products. It's the patience that we need to exhibit as we wait for these things to take off. That's really the message of the call, I think.

All right gentlemen. Thank you so much for your time this morning.

Thanks, Jeff.

We have no further questions at this time. I will now turn the call back over to Mr. Perham for any closing remarks.

Yes, I do have a few closing remarks this morning. And this call today, I was listening to myself and I think I want to touch on a couple things.

I mentioned that the activity level that potential customers, and through the various regions that our sales force lives in is at an all-time high. And I just would like to -- and it is. And I would like to say that customers are beginning to appreciate the power and the capability of our increasingly smarter family of sophisticated networking memories. And as we move up the performance ladder, and we're moving fairly swiftly because of the enormous demands on the Internet, I'm feeling very good about that.

And then the second point was, and this gentleman just got me to talk about that, so I won't say much, I continue to anticipate additional design wins. And I'm going to find a way to make it known. I haven't figured out, Jim, yet how I'm going to do this.

But I said something would happen in the fourth quarter. I mentioned that last quarter and we're in the fourth quarter. And we still are optimistic that we can do something and lock down a few more in this quarter.

We talked about the attributes for -- and I won't go into it much more, other than to say BE2 allows substantial performance upgrades for guys who just need to run faster. It's a substantially improved feature set to allow you to more effectively solve the new problems that we found via BE1 by working with customers. So we're solving new problems and doing a better job of solving old ones on the line cards and network interface cards.

The power of our new higher speed more feature-rich solution has stayed relatively flat, so we didn't put more heat in the guy's box. And the cost basis of this part will allow us to put our customer on some kind of sensible price learning curve.

You know, Jim mentioned we scaled the Company back a little bit in this quarter and I'd just like to say that I don't like to do that, and it's a tough thing to do. But we have to bring the Company into some kind of a configuration that allows us to serve the business which -- that we want to be in, which is fabless semiconductor Company. So we've addressed cost across all fronts.

And the last thing that I think I didn't make clear enough on my -- in my prepared remarks was that I have -- that I don't believe businesses make sense until they make money. And I'm personally and totally devoted to the concept of driving substantial value to the shareholders and to the employees who, for the most part, are one and the same. And at the same time bringing a continuous stream of solutions to the customer that makes you a very, very important supplier to him.

And that's really what it's all about with me. So that's all I have for this morning. I really appreciate the time and I look forward to talking to you all again soon. Thank you very much.

Thank you for your participation in today's conference. This concludes the presentation. Everyone may now disconnect and have a great day.