Rambus Inc (RMBS) 2003 Q3 法說會逐字稿

Good afternoon. My name is Lasandra , and I will be your conference facilitator. At this time, I would like to welcome everyone to the Rambus quarterly earnings conference call. (OPERATOR INSTRUCTIONS). Mr. Bob Eulau, you may begin your conference.

Thank you, operator. And welcome everyone to the quarterly conference call from Rambus Inc. This will cover the financial results of our third quarter in 2003. My name is Bob Eulau, and I am Rambus' Chief Financial Officer. Geoff Tate, our CEO is joining me today.

I will start with a summary and analysis of the financial results for the quarter, and an update on litigation. And then I will turn it over to Geoff for a discussion on the business progress we have made in the last quarter. We will then open the lines for Q&A, and both Geoff and I will be available for questions.

The press release has been filed with the SEC on Form 8-K. If you have not received it, it is also available on our website at www.Rambus.com. A replay of this conference call will be available for the next week at 800-642-1687. You'll need the ID number, 3283064. In addition, we are simultaneously webcasting this call and it can be accessed on our website beginning at 5:00 Pacific daylight time.

Before we begin, I need to state that our discussion will contain forward-looking statements regarding our financial prospects, including revenues, costs and expenses, net interest and other income, and tax rates, pending litigation, demand for our products, development plans, anticipated product production and shipment dates, relations with licensees and other third parties, capabilities and technical specification of our products, and various other matters, and that actual results may differ materially.

Among the reasons which could cause actual results to differ materially is the possibility of inadequate shipments of RDRAM memory devices and controllers for the Sony PlayStation 2 in PC main memory market; the market response to these products; any deterioration in DRAM market, including declining prices; any delay in the development of Rambus-based products by licensees; any delay in the development and shipment of new Rambus products; a greater than expected response of the market to competing technology; a lack of progress on price and cost reductions by RDRAM suppliers; inadequate progress on signing new contracts for RDRAM, XDR, SDRAM or DDR memory devices and controllers; inadequate progress on signing new contracts for RaSer and Redwood interfaces; current licensees not fulfilling their contract obligations; unexpected delays in completing contract requirements; current licensees terminating their contracts; adverse litigation decisions and other matters, which are available in our SEC filings, including our 10-K and 10-Qs.

Once again, Rambus delivered solid financial results for the quarter. For the third quarter our year-over-year revenue growth was very good at 17 percent. Our net margin increased slightly from last quarter because our litigation spending declined. The Company also made significant strides in our efforts to advance our technology.

And finally, we made progress in our ongoing efforts to defend our intellectual property. And as Geoff will discuss, we made very solid progress in several areas of our business.

Now for a bit more detail on revenue. Total revenue for the quarter was $28.6 million, up 17 percent over the same period last year, and sequentially down 2 percent. Total royalties for the quarter were $24.6 million, which is up 6 percent from the same period last year, and down 4 percent relative to last quarter. Please remember that this quarter's royalties reflect shipments that occurred in the second quarter of 2003.

SDRAM and DDR royalties were up sequentially, approximately 1 percent for memory devices and controllers combined. To help with scenario planning, the revenue specifically coming from SDRAM and DDR memory, memory controller royalties comprised about 41 percent of our revenue for the September quarter. Overall, our RDRAM interface royalties were down 38 percent from the September quarter last year, and down 27 percent from last quarter. We believe this royalty decrease reflects the decline in shipments of PCs using Intel's 850E chipset.

Contract revenue in the third quarter of 2003 was $4 million, which is up 222 percent over the same period last year, and up 7 percent from last quarter. This increase in contract revenue reflects the revenue that we are recognizing for the XDR memory interface and Redwood contracts. We're pleased by the steady adoption of our latest technologies. We received about $3.1 million in cash on contract payments this quarter. And most of revenue recognition for this cash has been deferred.

Total costs and expenses for the third quarter were $22.2 million. This was down 1.7 million, or 7 percent, from the previous quarter, and this was up 5.2 million, or 31 percent, from the same quarter a year ago. Cost of litigation was down 1.7 million from last quarter at 4.7 million, and up 1.9 million when compared to the same quarter last year.

While litigation costs were down from the last two quarters, they remained high primarily due to the case with the Federal Trade Commission. The FTC hearing continued through July and August with substantial briefing in September. I will provide more detail on this and other litigation in a couple of minutes.

Operating expenses, excluding cost of litigation, were flat when compared with last quarter. And they were up 23 percent when compared with the September quarter last year, with the increase largely attributable to our planned ramp up for XDR Redwood and RaSer interfaces.

From a profitability standpoint, this was another solid quarter. Net income for the third quarter was $5 million or 18 percent of revenue, compared to 4.5 million in the previous quarter, and 5.9 million in the same period last year. This continued profitability led to good cash flow for the quarter. Our cash, cash equivalents, marketable securities and restricted investments during the third quarter were 182 million versus 191 million last quarter, and 188 million as of December 2002.

While this number is down $9 million from last quarter, we invested almost 15.1 million in repurchasing shares of our common stock. The share repurchase program is an ongoing long-term program. Repurchases for the September quarter totaled just over 800,000 shares. And our average purchase price for the quarter was $18.70. We have a remaining authorization from the Board to repurchase up to an additional 3.8 million shares over an undefined period of time. Volumes of purchases will vary by quarter based on the opportunity. At the end of this quarter, we had average basic shares outstanding of 97 million, and diluted shares outstanding of 106 million.

Now I will give you some guidance on what to expect in the fourth quarter of 2003. This guidance obviously has a lot of uncertainty associated with it. This guidance reflects our best guess at this point in time, and our actual results could differ materially from what I'm about to review.

We estimate that revenue will be up slightly to between 29 and $31 million. We estimate that our operating costs and expenses will be between 21 and $25 million. Based on what we know today, we believe litigation spending in Q4 will represent somewhere between 3 and $6 million of our operating expenses. The increase in spending is primarily driven by increased investments in our XDR memory interface in Redwood product areas. Litigation spending is always difficult to predict, because we do not control the timelines and requests from the courts, nor do we control the actions that our adversaries may take which cause us to incur more expense. Evidence that is produced in the hearings can also cause us to change our plans.

Where we end up in our estimated range of 3 million to 6 million is very dependent on how quickly we go to trial in Virginia, and the activity level for other private litigation. We're anticipating net interest and other income to be between 900,000 and $1 million. And finally, we're still estimating a tax rate of 32 percent for 2003.

On the subject of litigation, I will review the major items in our litigation timeline. Let me start with an update in the U.S. As we reported last quarter, Infineon had sought for the review by the U.S. Supreme Court of the rulings in our favorite entered last January. Infineon filed its petition for on July 3rd, and its petition was denied on October 6. This means that the broad reading of our patents by the federal circuit remains the law of the case. It also means that Infineon will no longer be able to assert the counterclaims rejected by the federal circuit.

We expect that the federal circuit has or will soon issue the mandate which will start the process of a retrial of our infringement claims against Infineon in the Virginia court. Our infringement claims will be tried together with any claims or defenses that Infineon has that are found to have survived the ongoing appeal.

Currently the date of the trial in Virginia has not been set. And we think our first conference call with the court regarding the Virginia case will be sometime late next week. Preparation for the anticipated Virginia trial will affect our fourth quarter expenses. One interim fixed royalty agreement with a DRAM manufacturer remains in place pending further litigation progress.

In private litigation against Kinex, discovery is ongoing with a trial date currently set for November 2004, and a Markman hearing, which is currently set for March 2004. In the private litigation against Micron there is nothing currently on the court's calendar.

The other major action recently in the U.S. has been with the Federal Trade Commission. Closing arguments were completed on October 8th, and one would typically expect the Administrative Law Judge to render his opinion by the end of the year. The next step will be to have the Commissioners of the FTC review the case and vote on whether to adopt the findings of the Administrative Law Judge. It is expected that the full Commission could take up to a year or more to render its decision on the recommendations of the Administrative Law Judge. If the FTC attempts to enforce a remedy, then Rambus will have the right to appeal to a Circuit Court of Appeals in the U.S.

In Europe the Mannheim Court has currently scheduled the next hearings in the Infineon and Micron cases in Germany for May of 2004. We're waiting for the European patent office to schedule the appeal hearings regarding the validity of our European patent, but we expect this to occur toward the beginning of next year.

All of the foregoing litigation items relate to Rambus' efforts to protect its intellectual property and to answer related claims concerning our 1991 to 1995 participation in the JEDEC Standard Body organization.

As I mentioned last quarter, we are now evaluating what antitrust or other remedies Rambus may pursue based on evidence indicating that there were concerted efforts, beginning in the '90s, to deter innovation in the DRAM market, and to deter market acceptance of Rambus' RDRAM product. We are not, at this point, prepared to detail this area further. If we elect to bring such claims, we can expect our litigation expenses to be more towards the high end of the range of projections that I gave you.

One final area that I would like to discuss is insider trading policies and activity. Rambus has an objective to ensure that insiders are motivated to work towards the financial success of the Company, and at the same time comply with applicable insider trading laws. We also recognize that insiders will choose to diversify their holdings from time to time.

Last quarter the Board approved a change in our insider trading policy that allows Directors and Officers to adopt what are commonly known as 10b51 trading plans. By entering a plan and a time when he or she has no material nonpublic information, an insider gives up the right to control the timing of sales of stock, and within certain parameters, agrees to a pattern of recurring sales for a defined period of time.

Two directors, Mark Horowitz, one of our cofounders, and Geoff Tate, our CEO, adopted these plans last quarter. And within broad constraints, you will begin to see periodic sales of some of their holdings pursuant to these plans. In the future, other Officers and Directors may choose to adopt similar plans.

That covers the key financial and legal issues for the third quarter of 2003. So now I will turn it over to Geoff for some comments on our business.

Thanks, Bob. My focus is going to be on operations. I would like to say first on progress. The operations hit a busy quarter of our product line to give you some update.

First, in the RaSer product line -- sorry, I had to reposition my microphone here. I hope you can hear me better. In the RaSer product line we've had significant improvement in momentum. We signed over 20 customer contracts to date. Our RaSer product line pipeline consists of 1 to 10 Gbps serial links for applications, such as Serial ATA, Zowie, InfiniBand and other applications.

And the topic I would like to focus on today is PCI Express. The PCI Express market we think will be significant, and we're working hard to be well prepared for it. We have signed a contract with ALi, who is integrating our PCI Express 5 into their new chipset, their core logic PC chipset.

\And we are experiencing significant interest from other companies as well. We're sampling PCI Express 5s to our customers on C's 0.13 micron processor technology. And have a roadmap for having our 5s available on all of the major popular foundries and processors. We have signed partnership agreements with two key partners to help accelerate our penetration in this marketplace, Synopsys and Denali

With Synopsys, we will be demonstrating a joint PCI Express compatibility platform at the PCI Express interoperability workshop. With Denali, we also are working on a joint design and verification platform. So by working together with these companies, combining their act technology with our 5 technology, we're providing a total solution to our customers.

In the Redwood area, we had announced in the past our plans for a high-speed parallel bus, which we have licensed to some customers already. And I pleased to say that this past Monday we jointly announced with Toshiba Corporation and demonstrated on Toshiba's evaluation chip the Redwood interface technology running at speeds up to 6.4 GHz. This is in Toshiba's 19 nanometer ASIC process.

We're very pleased with the progress and achievement of this technology milestone. And we have been busy demonstrating this technology to dozens of companies, including at the recent Intel Developers Forum.

On our XDR memory product line, we talked last quarter at the Rambus Developers Forum in Japan, and detailed significant aspects of our plans. When we recently announced our plan to extend the XDR memory technology to PC main memory. And we anticipate in the 2006 timeframe that this level of performance will be required by high-performance PCs.

The ingredients necessary to bring XDR to PC main memory include a broad range of memory modules, which we called XDIMM's, programmable with XDR DRAM's, buffers, connectors, clock generators. And a comprehensive system design, set of documentation and guidelines, all of which we're working on now. The XDIMM memory module will provide over 12 GB per second of bandwidth at 3.2 GHz on a 32-bit bus. This is four times the module bandwidth of DDR2 DIMMs at 400 MHz. It will enable us the highest performance and lowest cost PC main memory systems.

We also unveiled other aspects of the technology required for PC main memory, like our dynamic link system topology. XDR also has a roadmap going to 6.4 GHz and beyond, which will allow us to further scale performance. And we can also increase the widths of these devices and the modules in the future as required, enabling in the long-term memory system bandwidths over 100 GB per second, 16 times more than today's 6 GB per second DDR main memory systems. That is the maximum you can get out of DDR today, as I understand it.

XDR DRAMs can be manufactured in multiple speed DIMMs, device densities, and device widths, with densities ranging from 256 MB to in the future 8 GB, and device widths ranging from by x1's to x32's. XDR DRAM's as a product line will satisfy the needs of both high bandwidth and high-capacity systems. We expect the first samples of XDR DRAM devices in 2004 with production ramping in early 2005. And we're working with our partners, Sony, Toshiba, Elpida and Samsung to achieve this, along with support from Cyprus ICS and TSMC who are working on the clock generators.

Finally on the RDRAM front, Sony recently announced that 60 million PS/2 systems are being shipped, which I believe you're all aware, use two RDRAM devices each. As well, Samsung has recently announced its move to 1200 MHz RDRAM in mass production, which we expect will be utilized in several applications, including the SIS four channel chipset for PC applications, which is anticipated to be officially unveiled in November.

The SIS R659 chipset will drive four channels at 1200 MHz RDRAM memory, with a peak bandwidth of 9.6 GB per second. This represents a 50 percent higher memory bandwidth than competing dual channel DDR chipsets.

Finally in the marketing area, I would like to welcome Ron May to the Company. He joined us in September as our new Vice President of Corporate Marketing. Ron joins us from Synopsys where he was Vice President of Corporate Marketing and Communications. Ron will be responsible for communicating our corporate strategy and plans to a wide audience, including through public relations events, the web site, and advertising.

Rambus has been, and will continue to participate in a number of industry events. We had a strong presence at the Intel Developer Forum in September where we showcased our XDR memory interface and our PCI Express applications. This week we participated in the Microprocessor Forum, where the first time we publicly demonstrated the Redwood parallel bus interface, as I mentioned. Additional conferences we participated in over the last few months have been the Communications Design Conference, the Network Processor Conference, and Denali's MEMcon conference.

In addition, to participating at industry conferences to carry our message to design engineers across the United States and the world, we have also begun hosting a number of U.S. and Canada regional design seminars. Our intention here has been to follow up on our success of our RDF Rambus Developers Forum in Japan, where we announced last quarter we had hundreds of engineers attend.

In the U.S. and Canada, we have decided to take a different approach, and instead of having a single conference, have regional design seminars going closer to the engineers to, hopefully, reach more of them. We have begun this design seminars series, and to date we're very pleased with results. We had almost 100 people in San Jose, about 50 in Austin, close to 100 in Ottawa, Canada. And in Southern California, we already have a couple dozen people signed up.

In addition, we have held our first Web seminar, and we had well over 100 people listen in on that. So we're pleased with the initial results and the interest in our products through our regional design seminars, and plan to extend and continue this program.

So it has been a good quarter, a busy quarter with significant milestones achieved with our customers, and on our products and technology development. And we're looking forward to continuing in busy Q4. At this time, Bob and I will take your questions.

(OPERATOR INSTRUCTIONS) Mike Crawford, B. Riley & Co.

Are you still expecting to see the first XDR sample from Toshiba next quarter?

The exact schedule by quarter, we haven't indicated, but we would expect the first samples to be in the earlier part of next year.

I think at the RDF in Japan, Nokita, Samsung and Toshiba all indicated that they would have samples early next year.

And of the 20 RaSer contracts, which ones can you disclose, or with whom?

I think that we have disclosed through press releases or other means any of those contracts, which our customers have agreed to let us disclose. The majority of them, our customers, would rather keep confidential at this time.

And another question relates to the interim agreement with a licensee that is paying at a reduced fixed rate. What needs to occur for that to revert back to the agreed upon license rate?

That is actually a common question right now. And the reality is we have not publicly stated exactly what the trigger is for that. And at this point, we don't think it is an our shareholders' best interest to do that.

And the final questions relate to the Administrative Law Judge's questions at the end of the closing oral arguments at the FTC. And you can read in the trial transcript that he was asking why weren't other participants punished for not disclosing patents or applications. And in one case, the complaint counsel said that Micron, in fact, did disclose a patent some 16 months after the fact, but they did so voluntarily. And I can read that in the transcript and I'm wondering what was the sense in the court when the complaint counsel cited that disclosure as voluntary?

Mike, it is John Danforth, can you hear me okay? Actually I would just say that the point that was being made there about others not disclosing wasn't, I think, a point the judge was making that there were other violations. The way he raised it was he said how -- he said this to complaint council -- how do I square your interpretation of the rules with the fact that so many others were not complying with what you say the rules were?

Well, he wasn't playing an accusatory finger. He was saying, look, all of this evidence seems to square with Rambus' reading of the rule, not with complaint council's. With respect to what the reaction of the court room was, there was a moment where complaint council said, gee, your honor, you shouldn't blame Micron for this because after all they voluntarily made the disclosure 15 months later. And there was a great deal of laughter in the court room because, of course, our whole point has been that the disclosure obligations, as written by Gentex , are written as voluntary obligations. A few sentences later complaint council tried to correct that.

And then the last one relates to the judge's questioning about -- he did say that there is evidence that there was some potential concerted effort to suppress the adoption of RDRAM. And that one of the reasons that it was cited that RDRAM failed was because of the high cost. But then he turns around and says that DDR has all the features in dispute here on RDRAM. How come that doesn't cost high? I was just wondering what you thought of that exchange?

Well, the interesting part about that exchange is that he is focusing on the evidence of collusion, which has not been a focus of ours in this trial. It is relevant we think to several issues, but if you read the transcript, it is not something we spent time with in the hearing. So it is interesting that he took the initiative to raise that.

Jeff Shriner , MS Capital Management.

John, can you choose the circuit court if the FTC were to rule against Rambus?

Yes, we can go to any circuit court we want, except the federal circuit has limited jurisdiction. And although we have not concluded our analysis, many people think it does not have jurisdiction that would cover the FTC.

And, Bob, could you clarify on the operating expense guidance you gave? The 21 to 25 million, did that include the litigation expense of 3 to 6 million?

Yes, it does include the litigation expense.

Then in terms of looking forward here, could we possibly see an increase sequentially in terms of RDRAM as a percentage of revenues in Q1 '04 due to the release of the SIS 659?

Well, we've got multiple factors at play here. One is we know that Intel is not producing the 850E anymore, and that is clearly depressing sales of RDRAM from what they once were. The 659 chipset will be helpful. And I also think that seasonally it is a fairly good period on the consumer marketplace where we have had some good success.

Can you talk about any impact the Company may have already seen, or looking out in terms of this new standard-setting body that has come out in the 10 GB field, the UXPI, and how they specifically said they want to exclude any Rambus technology?

All we know there is what we read in the papers. But we think that we have an excellent technology. And we're working with numerous customers who are interested in our 5 to 10 GB per second technology. And we're hearing positive feedback from them that we're one of the few companies, or maybe the only company, that has done our homework there. So it is hard to tell. It is early days. But we're pretty happy with our competitive position at this point, and hope it will work out well for us.

And, Geoff, can you make any comments in terms of if there are any products, and if so how many, are currently shipping that are using the RaSer technology?

Today there are no products to my knowledge shipping using the RaSer technology. I think that we have seen some products demonstrated. Internet Machines, as I understand, demonstrated their product. Exactly where they are at on shipping, I'm not quite sure. But it usually is a couple of months from demonstration to shipments.

So I think we're getting close. And we've got other customers in the pipeline. But it is not as far along as we would have hoped in the past.

And one last question. Is the Company still on track for the 400 patents at the end of this year?

400 patents at the end of this year. I don't remember us saying that was the track we were on. But we have shown a slide before in our investor presentation where we have issued patents, I think, somewhere between 200 and 250. And we expect to have a total of around 500 filed and issued patents by the end of year. And we're making good progress. But that is a combination of patent applications and issued patents. (multiple speakers).

Okay. I apologize for incorrectly asking that one.

No problem.

In terms of the segments, what would the majority of these patents, or the patent applications more say, be covering? Would it be more related to RaSer or XDR?

Well, I think that people really have to go and read the patents and determine for themselves. And there's a lot of commonality between our product lines as well. All of our technologies that we have XDR, RDRAM, RaSer and Redwood involve high-performance chip-to-chip I/O, and that underlies, and is core technology that is shared among all of them. The exact expression and implementation can vary from product to product. And then we have architectural inventions as well that go into the product lines. So XDR and RaSer have a lot of commonalities, and there could be some patents and inventions we have that are using both.

Thank you very much, gentlemen. Good quarter.

Garo Toomajanian, RBC Capital Markets.

Geoff, you expressed a lot of enthusiasm for RaSer, and in particular for PCI Express. I am wondering if you could quantify what that opportunity might look for Rambus?

And then after that as a follow-up maybe talk about what the competitive environment looks like now, where sort of the sweet spot is for you guys to win in that?

Well, of course, I reiterate that we don't have a crystal ball, but we see that Intel is showing a lot of enthusiasm for PCI Express, and is expecting it to be used very widely in their product lines. And we see a lot of other companies planning to adopt Express and use in their's as well. So we have seen a lot of customer interest. We have had a lot of customers talk to us and say that they are planning to use PCI Express. That is the reason for our enthusiasm.

In terms of our competitive position, we aren't the only people planning to offer PCI Express 5, but the feedback that we're getting is that we are -- I don't want to say -- use the term in the lead -- but we're certainly one of the leading people in the PCI Express 5 space at this point. And we've got a very good engineering team. We had a lot of experience doing high-speed signaling. To us, 2.5 GHz, which is the speed of PCI Express, is relatively slow. And I think that most of our competitors, it is their leading edge.

And it is not just a circuit design game, it is also very much signal integrity and system design. And I think that is where we really excel compared to our competitors. So we feel like we're well-positioned, but it is the competitive game, and we will see how it plays out.

(OPERATOR INSTRUCTIONS) At this time, there are no further questions.

Okay. Well, what I would like to do is thank everybody for listening in today, and thank you for the questions. I want to remind you that we are planning to have an analyst day. This is going to be November 12th in New York. And we will be webcasting our analyst meeting for all our shareholders.

You might also be interested to know that we have been invited by NASDAQ to open the market that day. So we're looking forward to having a good meeting. And I would like to thank everybody again for joining us today.

Thanks everyone.

Thank you for participating in today's teleconference. You may now disconnect.