Rambus Inc (RMBS) 2007 Q1 法說會逐字稿

Good day, everyone, and welcome to this Rambus first-quarter 2007 revenue conference call. Today's call is being recorded. At this time, I would like to turn the call over to Mr. Satish Rishi. Please go ahead, sir.

Thank you, operator, and welcome to Rambus' first-quarter 2007 conference call. I am Satish Rishi, Chief Financial Officer, and with me today are Harold Hughes, our President and CEO, and Tom Lavelle, Senior Vice President and General Counsel. We also have Sharon Holt, Senior VP of Sales, Licensing and Marketing, with us for Q&A.

The press release for the results that will be discussed here today has been filed with the SEC on Form 8-K. If you want a copy of the release, please visit our website at www.rambus.com on the investor relations page under financial releases.

A replay of this conference call will be available for the next week at 888-203-1112. You can hear the replay by dialing the toll-free number and then entering ID number 6011140 when you hear the prompt. In addition, we are simultaneously webcasting this call, and a replay can be accessed on our website beginning today at 5:00 PM Pacific Time.

Before we begin, I need to advise you that the discussion today will contain forward-looking statements regarding our financial prospects, pending litigation and demand for our products, among other things. These statements are subject to risks and uncertainties which are more fully described in the documents that we file with the SEC, including our 8-K's, 10-Q's and 10-K's, and these forward-looking statements may differ materially from our actual results. In addition, the revenue numbers we present today through our press release and this discussion are not yet audited, and will be finalized when we file our financial statements.

Now, I will turn the call over to Harold.

Thanks, Satish, and good afternoon, everyone. In Q1, we turned in another strong revenue result of $50.1 million. Satish will go further into the numbers, but I wanted to comment briefly on our ongoing restatement.

We were all anxious to get through the restatement and back to reporting our full financials. I share the frustration of our shareholders, and Satish will provide an update on the restatement and investigation later in the call. The restatement, while an extremely important activity and a tremendously difficult one, as well as the FTC's situation that Tom will address, are issues that focus on the past. What I want to focus on today is the future. Specifically, I want to talk about the future of computing and that's very positive implications for Rambus.

We have seen a great deal of industry discussion around the topic of multicore processors. In fact, with the semiconductor industry's ability to continue to drive transistor scaling down to 45 nanometers, 32 nanometers and even smaller, we can pack an incredible number of devices on a single chip. In order to continue to scale performance and take advantage of all these transistors, the processor architects have moved to multicore. This is an exciting area for Rambus, as the common need across these multicore platforms is a veritable thirst for greater memory bandwidth.

There are plenty of examples of multicore processors where the manufacturers have just [ganged up pop, ganged up] additional cores which do the same set of tasks. What is more interesting, and what we believe presages the future of computing, are the multicore processors which handle both general compute and graphic tasks with a single chip -- processors that need both capacity and high bandwidth, and there is only one high-volume memory solution that meets both of these requirements, and that is Rambus' XDR memory architecture.

XDR DRAM is shipping today in millions of units per month. High-volume production XDR devices -- again, available today -- can provide bandwidth of up to 8GB per second, and the XDR family has a roadmap of devices bandwidths of over 25GB per second. The XDR architecture can provide 4GB of capacity, and with an easily supported, buffered architecture, can scale to the kinds of capacities needed even by high-end servers. No other memory architecture comes even close to serving both these needs simultaneously.

Not only is XDR an obvious performance choice, it can be the lowest-cost solution as well. In high-end HDTVs, which can require as much as 6GB of bandwidth, a single XDR DRAM device can do the job of four DDR2 devices. From both the standpoint of bill of materials and reducing system complexity, XDR is the economical choice.

This is a message that is resonating with our customers. In every one of our sales geographies -- be it the US, Europe, Asia or Japan -- we have customer engagement activities for XDR in compute and consumer electronics, such as HDTV and set-top box applications. Some of these are at very early stages, but some are well-advanced. Of flagship application is the PlayStation 3, and it has already shipped in the millions. Every PlayStation 3 ships with four XDR DRAM devices. We will make you aware of other XDR-enabled products across computer and consumer electronics as they are released into the market.

Regardless of the specific application, all of our customers are in high-volume markets, and all demand assurance of supply. This is why our Qimonda XDR DRAM agreement is so important. When Qimonda has its XDR DRAM products in production, we will have the current number one, number three and number five DRAM suppliers shipping XDR.

Support for XDR DRAM from these top-tier DRAM suppliers provides the assurance of supply our customers require. As we have said before, what we have achieved with Qimonda is our goal for all unlicensed DRAM suppliers -- namely, to put in place a broad patent license agreement and engage through a technology agreement on products such as XDR DRAM that add value to our customers' products and to our customers' customers' products.

The XDR memory architecture is ready for multicore processing today. We foresaw the need for high-bandwidth, high-capacity memory solutions long ago. Our business model requires that we anticipate and invent solutions years ahead of the market need. That's why we hire and must retain the industry's best and the brightest. In five years' time, leading-edge graphic and gaming applications, applications that will be powered by multicore processors, will require up to a terabyte per second of bandwidth, and we are working on the technologies that will deliver that, make that possible today.

Along with the need for ever-increasing memory bandwidth, customers are increasingly concerned about the power usage for their products. We recently showcased our low-power initiative, targeted multigigabit per second platforms. Our engineering teams achieved a 3X power reduction when compared to traditional implementations, while still delivering high-bandwidth performance. At a recent IEEE conference, we demonstrated a test receiver that transferred more than 3,600 terabits of data utilizing the power of only two standard AA batteries.

Let me put that in context. 3,600 terabits of data is equivalent to transferring the content of nearly 100,000 DVDs, accomplished with just the energy provided by two small batteries. That's an accomplishment, and one that is recognized not only by IEEE Society, through the presentation of our findings at their conference, but by the academic community as well.

We continue to bring value to the industry through our innovations. This innovative spirit is in our DNA, and our teams continue to push the boundaries of what is possible.

One measure of this is our continued process on building our patent portfolio, and in the first quarter, we reached an important milestone, with the issuance of our 600th patent. We now have 620 issued patents worldwide and another 477 pending. Consider that it was less than a year ago that we announced we had reached the 500-patent milestone. These 600 patents equate to 1.5 patents per Rambus employee, a metric that places Rambus in the top tier of innovation-driven companies.

This continued high output of patents represents our continued commitment to fundamental core technology research and development. During 2006, we committed to a 4X increase in the resources focused on core technology, and we delivered on that commitment. The continued growth of our patent portfolio signals that our commitment of resources is bearing fruit.

So, to wrap up, performance matters in computing and consumer electronics, and XDR memory architecture is the best memory solution available, period. XDR provides both the bandwidth and capacity required by an increasingly powerful generation of multicore processors. But the race never ends, and the bandwidth requirements of tomorrow will continue their steady march forward. Our engineers and scientists push the boundaries of what is possible, even demonstrating industry-leading low-power technology. What the market will need in 5 to 10 years, we are working on today.

With that, I will turn the call over to Tom to give perspective on what happened regarding the FTC, in addition to other legal developments during the quarter. Tom?

Thanks, Harold, and good afternoon. As we've said before, the issues surrounding the FTC arise from matters that deal with the past. However, some of the clarifying points made by the FTC do affect our business moving forward, and we want to provide some perspective on that for you today.

First, as a review, on February 2nd of this year, the FTC issued its remedy decision setting maximum royalty rates we can charge for Jetta-compliant SDRAM and DDR SDRAM devices and controllers. The Commission's remedy did not extend to DDR2 SDRAM devices or controllers or any subsequent generations.

Subsequent to that, on March 19th, the FTC granted a partial stay of its remedy order. This stay pertains to Rambus' ability to collect royalties in excess of the FTC-imposed rates. The FTC has ordered that Rambus is permitted to collect royalty payments for use of our patented technologies during the period of the stay, in excess of the FTC-imposed maximum allowable royalty rates on SDRAM and DDR SDRAM products, so long as the funds above these FTC-imposed rates be placed into an escrow account.

The funds in the escrow account will be distributed in accordance with the decision of the Court of Appeals when that decision finally comes down. Any funds which go into escrow will not been included in our revenue unless we eventually win the appeal. More on the appeal in a moment.

As part of its order on the stay, the FTC also clarified that its remedy is intended to be forward-looking. Therefore, Rambus is not restricted from collecting royalties for the use of its technologies in the past, nor are we required to refund royalties that have already been paid.

Now, back to the appeal. On April 4th of this year, we filed a petition with the Court of Appeals for the District of Columbia Circuit, seeking both a review of the liability findings and of the remedy findings in the FTC matter. As we've said before, we believe the FTC, in both the liability findings and remedy, got it fundamentally wrong. The Commission chose to ignore its own Chief Administrative Law Judge, the key fact-finder in the case, and contradicts the Federal Circuit's findings on the same issue.

We remain disappointed with the FTC matter, and will provide further updates on the appeal as that case proceeds forward. The remaining portions of the FTC remedy order that were not stayed went into effect on April 12th of this year, and Rambus is in compliance with the remaining FTC order, of course.

Turning now to our other litigations, on April 4th of 2007, Judge Whyte in San Jose held a case management conference with all the parties involved in the Rambus-related cases in his court. This included Hynix, Samsung, Micron and Nanya. As a result of that conference, the Judge has combined portions of the cases into one, since these cases deal with essentially the same issues. The combined cases will address the defendants' allegations of improper conduct by Rambus. Our allegations of patent infringement by the remaining defendants will be tried sometime after the conduct allegations. You will recall that Hynix has already been found infringing.

Judge Whyte has tentatively set a trial date for the conduct issues in January 2008. There will no doubt be pretrial motions that will be filed in subsequent rulings, and we will continue to post information on our website to keep interested parties updated. We also expect a claims construction hearing on the remaining patent claims in early 2008.

Turning now to the joint boycott case in San Francisco, this is the complaint we filed in 2004 alleging the parties engaged in a joint boycott to eliminate RDRAM from the marketplace. A case management conference was held in this matter on April 12, 2007. The judge in that case, Judge Richard Kramer, determined that a discovery referee was needed in order to better manage the large number of discovery issues that are likely to our rise in the case. The parties were ordered to meet and confer on the appropriate referee.

Also, Judge Kramer determined that it is too early to set a formal trial date, but mentioned March 2008 as an aggressive target. He also ordered that depositions and discovery can proceed while we work with the discovery referee to mediate issues as they arise. These depositions may include certain witnesses who are incarcerated for their roles in the DRAM price-fixing prosecutions by the US Department of Justice.

With that, I will turn the call back to Satish to review the revenue for the quarter and provide an update on the independent investigation. Satish?

Thanks, Tom. As Harold mentioned, total $50.1 million were up 6% over the first quarter of last year, driven primarily by patent licensing deals that we signed in 2006, Sony PS3 royalties and recent technology deals, including Qimonda. It is worth noting that revenue in the first quarter of last year included $10 million royalties from the Intel license agreement.

Given that we're in the process of restating our financials, we do not have the complete income statement details to share with you. However, I can say it was another good quarter, reflected by revenue results. In addition, overall cash, defined as cash, cash equivalents and marketable securities, increased $54 million from the first quarter of 2006, ending the quarter at $445 million.

Let me provide you with a quick update on the status of our financial restatement. I know that many of you are anxious to get this behind us, and so are we. When this began, we did not expect this process to take as long as it has taken. We have found that the process is tedious and time-consuming.

As we mentioned in the last call, the independent investigation into the stock option granting process had been largely completed. Subsequent to that, we have had to internally determine the correct measurement dates, verify the correct accounting that existed at that point in time and then calculate how the charge would be allocated to future periods. We also verified other equity-related activities such as vesting of options during leaves of absences, vesting of options post terminations and the correct accounting for employees who became consultants and vice versa.

None of the accounting impacts from any of these related activities had been recorded in the past, so we first had to determine what the facts were, which is the more time-consuming part, and then record the correct accounting entries. In addition, as a part of the restatement process, we have opted to go through additional procedures to ensure that other areas that are entirely unrelated to stock options have also been accounted for correctly. We have not found any significant issues from any of these non-equity-related areas, but nevertheless we have had to go through the process of ensuring that we have reviewed all possible areas where there could have been an issue. Again, this is a time-consuming activity, and we are working through it as best as we can.

In October of last year, we mentioned that we expected the non-cash, pre-tax stock-based compensation charge in connection with the misstated stock options to be in excess of $200 million. As of now, having done quite a bit of work on this, we estimate that the charge for all option-related issues, including misstated options, accounting for vesting of the terminations, et cetera, will be approximately $190 million. Of this, approximately $170 million will be booked as a non-cash pre-tax charge for the periods from 1997 to 2005, and approximately $20 million will be recognized during 2006 and beyond as the older, mispriced options vest.

On February 7th, we had also reported that we received a letter from the NASDAQ Listing and Hearing Review Council informing us that the Council had called for a review and also stayed the NASDAQ Listing Qualifications Panel's January 19, 2007 decision. The Council gave us the opportunity to submit additional information for its consideration by the end of March 2007. In March, we provided the Council with a status update on the restatement process, and we requested that they grant us an additional extension to complete our restatement. The matter is currently under review by the Listing Council, and we expect that they will issue the decision sometime soon.

Under the NASDAQ marketplace rules, the Council can provide us with an additional extension that is the lesser of 60 days from the date of the decision or 180 days from the date of the decision made by the NASDAQ Listing Panel, which was January 19, 2007. While there can be no assurance that the Council will grant our request, we are hopeful that they will grant us a further extension, which will allow us to remain listed on NASDAQ while we complete this restatement process. We will provide additional updates on this process via 8-K and press releases as appropriate.

As Tom mentioned, the SEC order is in effect for patent royalties on SDRAM and DDR memory devices and controllers that are shipped after April 12, 2007. We have also mentioned that we have filed an appeal of the SEC's liability and remedy decisions. As Tom outlined, the SEC has ordered that any royalties in excess of MARR, or the maximum allowable royalty rate for these products, be put in escrow until a final decision is reached through the appeal process. We are in the process of talking to our customers about the mechanics of this, and also working through the overall impact of this decision on our revenue in the second half of the year.

At this time, based on our estimates, we expect that this may put up to $15 million of our revenue in the second half at risk. However, as Harold mentioned, we are working diligently to grow our business, and are seeing momentum in our XDR memory architecture engagements.

Let me turn to our forecast for the next quarter. As mentioned before, we do not expect the FTC ruling to impact our revenue until the third quarter. Based on our current expectations, we estimate that Q2 revenue will be between $45 million and $50 million.

We have also received a few questions regarding our shareholders' meeting. I just wanted to remind everybody that we cannot have a shareholders' meeting until we are current on our financials. Once we get current, we will provide more details about the annual shareholders' meeting.

That concludes our prepared remarks. Operator, would you please open the call for questions?

(OPERATOR INSTRUCTIONS). Dunham Winoto, WR Hambrecht.

I'm calling in for Daniel as well. If I can just ask you a couple of questions, first of all, can you talk a little bit about the potential opportunity for PS3 this year?

We cannot give you figures as to the potential business impact or revenue impact for us from the PS3. But certainly, I think we have been very public about the Rambus technologies that are incorporated in the PS3. I think if you take a look at what Sony has released publicly about the volumes they have already shipped and what they expect to ship for the rest of the year, it at least gives you some idea of the positive impact for Rambus.

In that case, then let me just change gears a little bit then in my second question. I think Satish and maybe Harold talked a little about the potential of getting delisted. So what are the key dates that we have to watch out for here?

Well, like I mentioned, we have sent in our information to the NASDAQ review council, and typically they meet once a month, and we are not privy to when they meet. But once they make the decision, they have -- within their right, they can give us up to 60 days from the date of the decision or up to 180 days from January 19th. Those are the two dates that they can provide us for the extension at this point in time.

150 or 180?

180.

(OPERATOR INSTRUCTIONS). Michael Cohen, Pacific American Securities.

Hi, Harold. You touched on the need for XDR memory in platforms that have both CPU and GPU cores. What I was wondering is, I was wondering, with the previous licensing agreements with AMD, now that AMD and ATI are combined, would ATI need a separate licensing agreement, or are they covered under the original AMD deal?

The current AMD, which is the merged company, is licensed on a forward-going basis. So the business that was formerly ATI on a forward-going basis is now covered under the existing agreement with AMD.

If you could refresh my memory, was there a variable component to that deal? Or was it a one-time paid-up fee of, I think, $200 million?

We didn't announce the specifics of the deal. We did announce that the total value of the deal is $75 million over a five-year period.

Similarly, would Intel need any additional agreements if they chose to combine CPUs and GPUs on a single die, or would their past licensing agreement that's now paid up (multiple speakers)?

To the extent that the inventions were covered by their paid-up license, obviously, no. But as I said, we continue to invent every day. They are developing the product as it goes forward. Presumably they will focus on the highest performance.

So I can't answer categorically one way or the other. I can simply say that over time, we will develop a patent portfolio that will require licensing.

You previously announced a systems-level licensing strategy, sometime back. I previously asked if you had sent out letters of infringement to systems manufacturers. At that time, the answer was no, and I was wondering, have letters been sent out now?

I doubt that we said yes or no or anything definitive in response to that question. We do not talk about the status of ongoing discussions with potential licensees.

Can you comment on whether that's still a viable strategy, or one that you plan to go forward as previously announced, or is that now off the table?

We continue to talk to a variety of potential licensees at all levels of the value chain.

My next question would be for Satish. I was wondering if you could give us a number on the amount of outstanding convertible debt and also the number of shares outstanding.

The amount for the convertible debt is $160 million.

Can you repeat that number?

$160 million, 1-6-0. That hasn't changed from our last public number. In terms of shares outstanding, we have not done that calculation, because we can only do that after we complete our restatement in terms of the option and the shares and so on and so forth, so that number we will let you know when we restate our financials. But again, the number shouldn't have changed very much from what you had initially.

I would use $100 million.

Roughly $100 million?

Yes. It certainly makes the math easier, doesn't it?

Yes. It's an even number. Is there anything that could change the number of shares outstanding with regard to the investigation?

There are some things that could change, but I probably wouldn't speculate at this point in time on what they might be.

I'm a believer that you are eventually going to file a DDR3 suit, and I was wondering in terms of the timing of when that -- it's already a standard. Would it have to be on the market? Could you comment on any timing of when such as suit would come, if it were to come?

No, I think we're not really going to comment on what is very speculative as to what we might or might not to do in the litigation arena, looking forward. I think I'd just least it at that.

Jeff Schreiner, American Technology Research.

I wanted to ask a question, and I apologize if this was just asked by Michael, but I was kicked off. But I'm wondering if you could provide a little bit of the reasoning or what Rambus believes the D.C. venue offers as a basis for its choice of the FTC appeal.

What it offers in terms of a choice?

Well, everything done legally is based on your perception of the positioning that you will receive. So what is Rambus' perception of why you chose the D.C. Circuit, when you had several options within the Circuit Court to choose? Do you feel that there's an expertise there at the level of the Judge relating to this type of technology?

I think there's expertise in a lot of the circuits. I think the principal reason really focuses on the fact that a number of the Jetta meetings took place in and around the Washington D.C. area during the time 11 years ago, almost 12 years ago, that Rambus participated. We thought that's a pretty important fact involved in our decision where to file. I would say it's probably the most significant fact.

Switching gears, Satish, I was just wondering if you could comment. I appreciate the initial guidance about the potential risk of revenue on the second half from the FTC ruling. Would you be able to just -- not specifically comment on which licensees, but could you comment on maybe the number of licensees this may, in fact, affect?

No, I wouldn't at this point in time, because it is complicated and we are still in the process of talking to many of our customers at this point in time.

[Does that mean] talking with the customers, trying to determine what the actual rate may be, if they are paying six payments at this point?

What the actual rate might be, what the actual products might be. There's a lot of different, complicated factors in there. So I think what we felt is that we should be -- if we feel there is some risk to the revenue, that there's some downside, we wanted to be upfront and let everybody know that there is some potential risk to the revenue. But we'll do our best to make it up in other ways, obviously.

Just one final question, and I appreciate your time today.

Sharon wanted to add one more comment.

I just thought I would add, each of our licenses is somewhat different in terms of how it's structured. So working through the ramifications of the FTC remedy order with each one of our licensees is very much of an individual process. There isn't kind of a blanket answer for each one.

Just the final question I wanted to ask -- maybe Sharon, it's for you or maybe Satish. Could investors look to the fact that the option restatement hasn't been completed? Also, we haven't seen during that period, really, the announcement of any new licensees. So could investors assume that when management is able to finally work its way through this process, that there could be the potential that that opens up the ability for the Company to possibly start to craft negotiations that may be in process, let's say, or in the works right now?

Well, I think the two tasks are unrelated. I think I'm focused quite a bit on the stock option investigation and the restatement, and Sharon and her team are focused outwardly with the customers and licensees. So yes, I would like to be spending more of my time on the business side of things, but it doesn't mean that Sharon and her team and the rest of the Company are not doing that. So I think the two tasks are fairly unrelated.

I think also, if you're wondering, are we holding back on any announcements or sitting on any announcements during this time, the answer is no. We continue to try to move new opportunities forward, as we always do. As we close new deals, we will announce those as an appropriate, at the time.

Hamed Khorsand, BWS Financial.

My question was related to if there is a set timeline where we can see XDR in a broader line of electronics?

Certainly not a set timeline. I went into it in great detail, because I believe it's probably more accurate to say that we increased significantly the likelihood that it will happen, based upon the architectural structure of these multicore processors. I obviously couldn't commit to a particular timeline. I'll ask Sharon to commit to a particular timeline, but I could never commit to a particular timeline.

I think, just to add onto that, certainly, as we mentioned earlier in the call, the addition of Qimonda as a third supplier of XDR in the market really does help fuel momentum for XDR in more applications. We are obviously trying to leverage that as we go out and talk to many more customers this year about XDR in a broader variety of markets and applications. In particular, we are very excited about, in the digital consumer area, the HDTV and set-top box market, which really is a perfect application space for XDR. As Harold mentioned, we believe longer-term, with the advent of multicore architectures in computing, that we will see XDR penetration into the computing market as well, and we are pushing with everything we've got on that this year.

In terms of getting into the Flash market with Spansion, is there a set product that Rambus is looking to attract with the Spansion deal, or are you going to expand on that arrangement?

We're continuing to have dialogue with Spansion about new technologies and potential areas for collaboration, but nothing that we're prepared to talk about today.

This will conclude today's question-and-answer session. Mr. Rishi, I will turn the conference back over to you.

Thank you, operator. That concludes our conference call for the first quarter. Thank you all for your continued interest in Rambus. Thank you.

Ladies and gentlemen, this will conclude today's presentation. We do thank you for your participation, and you may disconnect at this time.