To: KJ. Moy who wrote (25180 ) 12/11/1999 10:51:00 AM From: J Fieb Read Replies (2) | Respond to of 29386
" To Infinband and BEYOND..........." All in all, from the relatively brief materials I've seen on the group's Web page, this is a good specification and a great move for the industry and for us. Perhaps the most interesting bit of news is that Microsoft is on the steering committee for the specification. Love it or hate it, Microsoft is the 800-pound gorilla that this group needed to dance with, and it's great to see the company as part of this effort. Also sponsoring the effort, but not on the steering committee, are the major networking hardware vendors, the chip vendors and Adaptec. It should be interesting to see how Cisco, Lucent, Nortel and 3Com use these specifications in their systems. About the only bad news to come from all this is the rather protracted time frame for the rollout of the specification and for subsequent systems: The draft specification is due at the end of this year, but don't expect to see systems based on InfiniBand until sometime in 2001. So while the systems won't make the fall lineup of server introductions, it's worth watching the trailers to see how the specification develops. The rewrite of the sequel is exactly what the industry needs. I'll give it two thumbs up, four stars, a picture of a guy standing on his chair clapping, four-and-a-half bags of popcorn...well, you get the idea. Yes! this guy is saying that these FC seats are only getting more valuable as the show progresses. Wonder what they will go for in '03 or '04? After a little reading people will want to pay a lot for a few good ANCR seats. Come play in the BAND. InfiniBand? Architecture is a channel-oriented, switched fabric, serial point-to-point link architecture aimed at meeting the growing needs of I/O reliability, scalability and performance on servers at a good value. InfiniBand Architecture introduces the use of an extremely efficient engine that is directly coupled to host memory which replaces shared buses with a fabric of switchable point-to-point links. This approach decouples the CPU from the I/O subsystem (as opposed to today's load/store memory-mapped I/O) and addresses the problems of reliability, scalability, modular packaging, performance and complexity. CPU communication with peripherals occurs asynchronously with the I/O channel engine being responsible for moving data to and from main memory and allowing the bus to act as a switch with point-to-point links capable of near linear scaling with CPU, memory and peripheral performance improvements. WHo wouldn't want to do biz with ANCR now to cover both SAN and Band? WIll this help ANCR take market share from BRCD? I think so!