Analysis on IT trends and competitive strategies, with emphasis on micro processors, computer systems and networks. Based on latest news, backed up with real data, this site intends to provide a true and realtime picture of the fast changing IT landscape. This journal strives to be accurate on facts and sharp on criticisms. You may email your opinion to sharikou@yahoo.com or post comments here, be cool and intelligent.
Freelance journalist on IT matters. Some of my writings have been published on online IT journals. Any original content on this journal is Copyrighted, but it's free for non-commercial use. Any Trademarks used on this site belong to their respective owners. Some of the pictures are links. If there is any issue with the content of this site, please email sharikou@yahoo.com .
As reported here, AMD and NetLogic will announce a design monday. I guess it will be a NetLogic co-processor for security stuff while Opteron chews network traffic data via 2 HTT links, total 16GB/s of bandwidth.
posted by Sharikou, Ph. D. @ 4/21/2006 04:38:00 PM
8 Comments:
another nail in Intels coffin. Some one at Intel needs to take the bull by the horns and get on the HyperTransport bandwagon like now! otherwise its definitely over for Intel.
these innovations will be the tip of the iceberg for speciallist accellerator CPU's
all driving big business servers, no room for struggling P3 wannabees?
There has got to be *some* reason why Intel refuses to use HyperTransport or a similar technology/architecture.
This decision to stay with what they have can't possibly be based simply on the stupidity and incompetance of Intel's management team? Theres no way!
If I'm not mistaking, HT is 'open', thus royalty free.
I agree. And yet, I expected this to be a nail in AMD's coffin too.
The future will probably setups with multiple giant FPGAs with one single CPU remaining as a control unit, especially in the supercomputing range, but also in common server business.
Here is another interesting news:
http://www.theregister.co.uk/2006/04/21/drc_fpga_module/
Only the non coherent Hypertransport for I/O is open. The cache coherent for IPC is not. AMD has just begun to license ccHT to these co-proc makers.
Thanks for the FPGA link, it supports my expectations. In short terms this means a clear advantage for Opteron, but in the long run it is is clear that less and less CPU bound computing power is required as specialized hardware is becoming cheaper (in fomrs of FPGA modules) und as efficient integrated into the computing system as the CPU itself.
This must result in a drop of classic CPU sales. Time will tell how well different CPU bakers can react to such challanges.
so the idea is that a general processor is just a framework to farm out specialised work to the FPGA?
i think AMD should quickly either merge or invest in this firm.
@dba:
Most people are maybe not interested in high-performance FPGA, possibly because they never heard of it. But many people are interested in high performance, no matter where it comes from.
Today FPGA often is of limited use because of bandwidth issues - PCI-X is simply not up to the task to feed such processing monsters. HT should solve this problem as well as problems of CPU<->FPGA communication.
With FPGA on an Opteron board, it should be easy to destroy every SuperPi result an Intel Conroe can deliver - and afterwards you reconfigure the FPGA to make a big ScienceMark washing up ;)
Post a Comment
<< Home