NEWS: February 05, 2001
Rambus Makes Its Case
Company president and VP talk to ElectronicNews about Rambus’ plans
By Steven Fyffe
Rambus Inc. is not sorry. Micron Technology, Hyundai Electronics Co. and Infineon Technologies have all accused Rambus of using monopoly tactics to try to take over the DRAM industry. They say Rambus (nasdaq: RMBS) stole ideas from open industry-standards committees, such as the Joint Electron Device Engineering Council (JEDEC), and then used them later to blackmail the industry for royalties on SDRAM and double data rate (DDR) memory.
There have also been some well-publicized hiccups as direct Rambus DRAM (RDRAM) has gone into production. Last year, Intel Corp. was forced into an embarrassing recall of 1 million bad motherboards equipped with memory translator hub chips that were designed to link RDRAM-oriented microprocessors to SDRAM.
But Rambus' business tactics have gotten the company where it is today, said Dave Mooring, Rambus president, in an exclusive interview with Electronic News last week. And with Intel's RDRAM-based Pentium 4 now ramping up, RDRAM is set to dominate the PC memory market, he said. Also present at the interview was Avo Kanadjian, vice president of worldwide marketing at Rambus, Mountain View, Calif.
Mooring : "The reason this industry, I refer to it as the RDRAM industry, has been able to do it is because of the business model that Rambus has taken. We went through early on and worked out the total solution. There are a lot of technical details, but a lot of it is just the aspect of facilitating and catalyzing cooperation between industry leaders.
"The fact that at 800MHz this stuff is all shipping in high volume well ahead of the schedule of the alternatives in lower volumes I think speaks very well of the business model. Hopefully, at some point this room, and maybe your readers, will come to appreciate that more."
ENews : You see the business model as a success, but it has rubbed a lot of your customers the wrong way. Can you explain the discrepancy?
Mooring : "My view is that in any industry's transition there are leaders and followers, and winners and losers, and those who are leading like it because they are the ones who reap the benefits through extra profits and extra market share with their customers. Those who are behind resist significantly, and it doesn't matter whether it's memory or some other thing, it always divides into those who have an advantage and those who don't. I think you are seeing that very clearly right now. Until those who are behind catch up, this controversy from that segment of the population will continue. And it's a comparatively small percentage of the population, but it's a very loud percentage.
"The closest analogy is how the communications vendors probably felt about Qualcomm Inc. Qualcomm invented CDMA a decade ago and had some rather controversial discussions with the handset manufacturers, but today Qualcomm is considered a very successful company. The whole handset industry relies on them and Qualcomm has got a market valuation of $50 billion, but they went through a very painful education process with their customer base, and I think Rambus is doing the same thing. The industry is just getting used to how to deal with the stratification of the semiconductor industry and the new layer of intellectual property suppliers."
ENews : Several of the bigger memory makers are resisting you now. Is that because they are lagging?
Mooring : "Look at how Samsung Semiconductor Inc. will do financially, leadership-wise and with the OEMs this year. They are going to do phenomenal, and Samsung's competitors, some of them won't like that."
Kanadjian : "It's clear to me that since the second half of last year that the market would be in a shortage by the end of the year, and a lot of the DRAM manufacturers were somewhat complacent about producing something that has lower yields and investing in the future. The ones who wanted to differentiate and who did their homework are reaping the rewards today. There's no doubt about it that if they had it to do all over again, a lot of suppliers would prepare better for this winter."
ENews : Historically, there seems to be room for only one major type of memory in the PC market. Is that memory going to be RDRAM?
Mooring : "If there's only one memory type, yeah. Over time, performance matters."
ENews : What about DDR?
Mooring : "DDR is in an awkward position in the desktop PC because it's not the cheapest, which is what the value section wants, and it's not the highest performance, which is what the performance section wants. So for the PC, it's kind of in no-man's land. I think that's why, besides the infrastructure, compatibility, pricing and availability issues, it's in an awkward position. It's a 'tweener.' It's stuck in the middle."
ENews : A lot of people feel it's unfair of you to use SDRAM and DDR patents as a bargaining chip — charging higher royalties for SDRAM and DDR to try to encourage people to make RDRAM. They say that is not the way you should go about trying to promote your own technology.
Mooring : "We're not doing that. We are just trying to get paid for what we invented. It's pretty simple. If you take the device that we announced in March of 1992 and you look at the features on it ... they ended up in the DDR devices. We can send you a datasheet from our 1992 RDRAM that was made available by Toshiba Corp. and these features are sitting there in the device."
ENews : Did JEDEC copy your invention in writing the DDR specs?
Mooring : "That's our claim, right? That there are pieces of the RDRAM that ended up in some of these vendors' DDR devices."
ENews : What about other synchronous interfaces on a logic chip or in a chipset? How broad are your patents in that area and who might possibly have to pay you royalties?
Kanadjian : "We've indicated to you that we are talking to controller companies as well as DRAM manufacturers, but we cannot divulge any specifics until we enter into agreements."
ENews : Will Intel have to pay?
Mooring : "I can just summarize by saying we are in negotiations with Intel and other companies at this point."
ENews : You have trod on some toes and there is some bad blood out there. If you could go back, what would you do differently?
Mooring : "Really, from a business-model, technology-development-partnering perspective, I wouldn't change anything. (I would have done more) market education, appreciation of what we are doing as an R&D hub and as a snowplow for the industry. Any time you try to run things five times faster and you do it two years before the alternatives, you are bound to have issues, and we should have trained the market on the value we were providing."
electronicnews.com
Ö¿Ö |
