SI
SI
discoversearch

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
RMBS 98.83+0.8%Dec 4 3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Jdaasoc who wrote (56477)10/4/2000 1:15:19 AM
From: Bilow  Read Replies (2) of 93625
 
Hi Jdaasoc; Re repeater latency... This is in addition to the nanosecond or two required for propagation delay down the wires.

By the way, even Intel doesn't suggest that their i840 is a great solution for a typical desktop system: The Intel® 840 chipset was specifically designed to meet the needs of high performance multi-processor systems.
developer.intel.com

Anyway, here's a link to the Intel notes on the RDRAM channel repeater:

Product Features
6 rclk round trip delay for RSL signals throught he MRH-R (page 7 of 52)
(Also see Max Memory support on page 9)
developer.intel.com

With PC800, an rclk is 1.25ns, so the latency penalty is 7.5ns, but for i840s going out with slower memory the latency is proportionately longer. This may seem like a lot of time, but actually, this amounts to under 2ns to get on and off the chip, each way. I think that's stunningly good performance for a repeater, but the result is comparable to the effect of registering DDR.

-- Carl
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext