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 : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: Bilow who wrote (29380)	9/12/1999 9:12:00 PM
From: Dan3	Respond to of 93625

Re: I am unable to understand your calculations.... ARGHH! You are absolutely right. I was reprising some numbers I had put up about a month and a half ago where, for the sake of being conservative, I had assumed a 99% cache hit rate or 1 in 100. It was pretty boneheaded to move to 95% and not go to 1 in 20. I'm sorry for any confusion. Everyone started using the 95% number so I went with it without adjusting the rest of the calculation. 95% is, of course, much worse for rambus, I was trying to give Rambus the benefit of the doubt. A month and a half ago, I was trying to raise the issue of performance effects poor latency would have by pointing out that number of instruction retirements that would be missed while a 1 gigahertz CPU waited after a cache miss. I used the very conservative hit rate of 99% to show that even under those circumstances that impact would be serious. I am sitting here trying to finish a program, and checking on the lists every so often, and popping off messages in a rush. Still no excuse for that one. Red-Faced, Dan

To: Bilow who wrote (29380)	9/12/1999 11:22:00 PM
From: Dan3	Read Replies (1) \| Respond to of 93625

Re: Latency Impact An issue that I've tried to get discussed on the thread is rambus vs. Virtual Channel DRAM. VC is very similar to SDRAM but uses static registers to provide better latency and overcome some of the inefficiencies of the SDRAM bus. It is also suitable for either standard or DDR. I think that the impact of reducing latency will be far greater than the benefit of streaming data that is provided by rambus's on chip interleave of banks. VC is also no royalty, only 3% more die area, and uses existing testers, etc. I think that this is why AMD decided to go with VC and DDR instead of rambus. VIA, ALI, and SIS are all supporting Virchual Channel in their chipsets that are sampling now for Q4 99 release. I'd be very interested in your opinion of Virtual Channel. Dan some benchmarks: necel.com necel.com Datasheet: necel.com