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 : The New QLogic (ANCR) -- Ignore unavailable to you. Want to Upgrade?

To: Craig Stevenson who wrote (14873)	3/4/1998 11:52:00 PM
From: George Dawson	Read Replies (1) \| Respond to of 29386

Craig, I read the same T11 notes. My first thought was that the ideal set up that would cover the node interconnects would be 2 Gig FC with 1 Gig FC as the I/O. I know what you are saying about proprietary clustering methods - but Compaq/Microsoft will be a huge force to deal with. The last information I have on ServerNet the I/O was 50 MByte/sec per data path with scaling through a 6 port router to 300 MByte/sec. They also have some claims for latency, but my data would suggest it is still the same order of magnitude as FC (300 ns per hop). Pfister states that the real advantage occurs only if ServerNet is directly connected to the CPU/memory units (main system bus). Although I could not find any recent diagrams on the Tandem or Compaq sites, Pfister suggests that the Intel based vendors (Compaq, Dell) that use ServerNet do not use it directly connected, but through standard I/O buses like an "intra-cluster LAN". He also has reservations about this proprietary system as a standard. George D.