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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Estephen who wrote (126959)	10/29/2000 5:52:58 PM
From: combjelly	Respond to of 1575740

"I have to take issue with this conclusion first it is totally untrue in regard to my friend as LSI." LSI does what on the desktop? Nothing. What does LSI do? Chips for set top boxes. Now set top boxes, if they ever get off the ground, would be a good market for DRDRAM. DRDRAM missed their window for the desktop. Sure, AMD is hiring Rambus engineers. But until you can actually provide a roadmap that has a Rambus based AMD product on it, I'll make a rational assumption that AMD is putting those engineers on a product with a future, LDT. If the P4 meets current expectations, DRDRAM will have no future on the desktop at all... Sure, DRDRAM dominates workstations as far as numbers. x86 dominates workstations in numbers and they are almost all Intel chips. The K6, Winchip and Cyrix just never made the inroads expected in the workstation market. Heck, I'll go one further, when Compaq ships the 21364 in a workstation, DRDRAM will be in the fastest workstation. And when AMD makes inroads into this market, the percentage of DRDRAM will drop. "In addition RMBS has agreements with close to 25% of the DRAM industry for royalties on SDRAM, DDR " This is wrong. The biggest fish so far, Hitachi, has transferred their DRAM to a joint venture with NEC. That joint venture doesn't seem to be breaking their neck to sign with Rambus. That leaves them with what? 5% maybe? "The latency problem seems to disappear when you mount the RDRAM chip directly on the processor." This is stupid. If the DRAM is directly on the chip, why waste transistors on DRDRAM? Or is he talking about stacking the chips? Still a stupid idea. "As one can see understanding the technical issues is not always that helpful." And managers don't have to understand anything about the products their companies make. Another stupid statement. If you don't understand much about the products you invest in, and you get burned, you have no one to blame but yourself. "However, Samsung's White Paper on the topic concludes that RDRAM is superior to other memory alternatives in latency." This white paper assumes only one RIMM on the channel and ignores the fact that the critical word comes last in the burst. While it might not make a difference in certain non-x86 architecture, it is critical (no pun intended) in x86. And, of course, if you have more than one RIMM, a bad situation only gets worse. Estephen, while I know this is pointless, try to actually learn something about what you are posting about. Show me an entry on an AMD roadmap that shows Rambus. Show me a server from a major OEM, not a screwdriver shop, that has Rambus memory. Heck, show me Rambus memory in a workstation that doesn't have an Intel processor.