Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: that_crazy_doug who wrote (15897)	10/25/2000 12:59:50 PM
From: milo_morai	Respond to of 275872

<font color-red>Entering The DDR Chipset Game -- The coming months will bring much posturing among the leading contenders in the DDR chipset market. After introducing its first double-data-rate chipsets, Acer Laboratories Inc. is gearing up for full production. ALi aims to capitalize on Intel Corp.'s approval of the use of third-party DDR chipsets to boost the performance of the company's existing processors, a policy analysts expect to be extended to a new 0.13-micron-shrink version of the Pentium III, slated to hit the market early next year. President Miin Wu discussed ALi's position in the DDR chipset race with EBN's Asia correspondent Faith Hung. EBN: How difficult was it for ALi to develop a DDR chipset, and what were the major hurdles? Wu: There are always challenges when you move to a new technology. DDR is a new architecture, so everybody's cautious. We were very careful when we started developing our DDR version. We even did motherboard integration on an unprecedented scale. Surprisingly, things worked out very well. We didn't find it as difficult as we expected to develop a DDR chipset. The development has gone quite smoothly. It's not hard for our motherboard partners to design their boards. And the memory availability and cost look right. So, we believe the transition this time should be very natural. Anyway, implementing DDR wouldn't be more difficult than upgrading PC100 to PC133. EBN: Does DDR for current Intel Pentium III processors kill off Intel's 920 and 840 chipsets supporting Direct Rambus DRAM? If Pentium 4 adopts DDR, what is the impact on Rambus? Wu: It would definitely be a big blow for Rambus. Rambus won't be able to compete with DDR in manufacturing cost or availability within the next two years. The performance of DDR-enabled chipsets has proven better than that of Rambus. Rambus is so expensive, but its performance isn't at all better. End users certainly wouldn't favor Rambus. So I think the 815 chipsets, which are being watched by everyone, will be Intel's key products next year. Nobody cares about the 920 or 840. EBN: Where do you see DDR taking off first, in high-end desktops, the value end of the market, or portables? Wu: DDR for high-end desktops will be the first to take off among those three. Some mechanical changes are necessary when it comes to DDR. Such changes really don't work for portable PCs. In addition, the major issue for portables is not performance, but power consumption. That's why it would come last. EBN: When will ALi begin full production of DDR chipsets? Will they be available at the same time for Intel processors and Advanced Micro Devices Inc.'s Athlon processor, or will one precede the other? Wu: The volume production will begin around the end of the fourth quarter. DDR chipsets for Athlon will be available a few weeks ahead of those for the Intel CPU. EBN: What are the market projections for DDR chipsets for Pentium and Athlon in the first half of 2001 and also for the full year? Wu: Many people are predicting DDR chipsets will probably account for 30% to 40% of the [DRAM] market in the first half of next year and anywhere from 20% to 25% for the full year. We tend to be optimistic about the outlook for DDR because your cost may rise a little bit, but your performance increases anywhere from 20% to 40%. It's a very substantial improvement. EBN: How much will DDR chipsets contribute to your company's sales? Wu: It's hard to say. Let's put it this way: The 1647 chipsets [for K7] and 1651 chipsets [for Pentium III] that we're focusing on can be used in both desktops and portable computers. So they're definitely our major products for next year. Their revenue won't exceed the total of our other products in the first half of 2001, though. The two products would have a better chance to generate more revenue beginning in the second half of that year. EBN: Where does the situation stand on gaining Intel's approval to make DDR chipsets for Intel Pentium 4 (Willamette) processors? If approval is not forthcoming, will ALi make DDR chipsets for Pentium 4 anyway? Wu: That's a sensitive question. I have no comment. ALi has two principles: We prefer not to talk about our future products, and we don't disclose our discussions with our partners. EBN: What do you think about Via's assertion that it will make Pentium 4 DDR chipsets with or without Intel's approval? Wu: It's been Via's approach to tell the media about what it plans to do in the future. Sometimes Via delivers the results it promised; sometimes it doesn't. This is not to pass judgment on Via as being either right or wrong, however. Intel recently settled a lawsuit with Via, which suggests that Intel has realized that open and fair competition is good for the industry and the individual players. So I think Intel will allow third-party chipset makers, including Via, to manufacture DDR for Pentium 4. EBN: What do you think about Intel's reported plans to develop a single-data-rate PC133 chipset for Pentium 4? Wu: Intel's move indicates that the company won't insist on Rambus any more, even though Intel hasn't given it up publicly. Intel is allowing alternatives by supporting PC133. EBN: Intel is developing its own DDR chipset, and Intel's partner, ServerWorks, is developing DDR for the next-generation Foster and McKinley server processors. Do you see any likelihood that this may be adapted for Pentium 4? Wu: You would have to ask Intel about that. I don't want to make comments for them. EBN: Did ALi consult with Micron Technology to use any technology from Micron's Samurai DDR chipset? Wu: No, we didn't. We used some technology from AMD, though. EBN: Do you have plans for the integrated graphics chipset using the Nvidia graphics core? Do you have any plans to use this version to support Athlon in a unified memory architecture (UMA) version? Wu: We have several graphics partners in different projects. Nvidia is one of them. But our policy doesn't allow me to reveal how we work together with our partners or when we might introduce new products. As for the second question, the best I can tell you is that unified memory architecture is used in every graphics product. Using UMA doesn't really save you lots of money-probably $8 to $10 at most. If you want great performance, you would need to buy high-performance memory. EBN: What is the upgrade path for DDR? Is ALi considering DDR-2 chipsets? Wu: The soonest we would start designing DDR-2 is 2001 or 2002, but by that time frame, the earliest that DDR-2 can enter commercial production would be 2003. It takes some time to move to DDR-2, you know. There have been some signs that DDR is a hit. Therefore, we have reason to believe DDR-2's platform won't go through a lot of changes. Right now, we don't have to worry about DDR-2. Once DDR turns out to be a big success, we'll surely consider DDR-2. It's a very natural move. A Month Old but good review to see how things panning out. techweb.com Milo