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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: jim kelley who wrote (68572)	11/15/1998 1:09:00 PM
From: rudedog	Read Replies (1) \| Respond to of 186894

Jim - an overall architecture rather than just an I/O architecture If you take a look at some of Bob Horst's SAN work or the other MESH-based architectures you will see that the architectures being proposed are comprehensive switch fabrics of which I/O is just one portion. Likewise memory and processor access are sub-components. These switch fabrics have capacity (measured) in the terrabyte per second range and latencies comparable to memory access thanks to wormhole routing. Point to point connection is not required for any operation let alone for the duration of a transfer. Memory semantics are supported but not required, the base paradigm is message oriented. You are looking at this from the perspective of current architecture, with all of the memory and data hierarchy baggage. That is not the way current development is going either on NGIO or in the competing technologies. Backwards compatibility requires special bus protocol converter chips. Not if the bus itself is supported as a component.