AMD Using HT For Co-Processors

a coprocessor would be a good way to get sPI times that beat conroe's :D
 
"I was fortunate enough to be able to ask specifically if we will see this technology integrated into AMD’s 4x4 enthusiast platform and the AM2 architecture. I was told that while there were no specific plans for HTX add-in co-processors in AM2 architecture at this time, there is the definite possibility should consumer demand crop up."

interesting
 
Potentially interesting, but it seems like it will be a long way off for gaming uses. Only time will tell.
 
Sorry, didn't ask that question specifically, but I did mention the AM2 socket specifically - his answer sounded like it encompassed AM2 as well. What was impressive was that AMD was focused on "customer centric" solutions. Many features and innovations that they were touting today came directly from requests from partners in the industry. They genuinely believe in listening to their customers. So I have a strong feeling that if technology like physics processing matures that they'll take this stuff very seriously. Having streamlined access to 3rd party specialized processors could be a competitive advantage for them, especially in the enthusiast realm. It could also help stave off "leap frog" tech jumps like Intel is capable of: case in point would be Conroe. If AMD could have an FPU or some other co-processor plugged in that assists where their technology is weak compared to next gen solutions, it would definitely make people take a second look as to which platform they'll buy or even if they feel like upgrading. (But hey, that could backfire - what if you could add in a 3rd party tech that essentially brings you up to current processing power with next gen platforms - why would you need to buy more AMD tech?)

For instance, I'm going to be going Conroe in all likelyhood for my next upgrade. If I currently had a board that had an HTX expansion socket or slot that I could pop in a media encoder processor, I'd probably just pop it in there instead of investing in a new platform...there's a lot of possibilities here and it all depends on market demand.
 
I peronally am still dreaming of a nice co-processor for speeding up raytracing and such - especially after mental images started to collaborate with ARTVPS in order to develope a unique graphics processor for photorealistic ray tracing... It would just be too sweet if I could plop one of these into a standard socket :D

(cf)Eclipse said:
a coprocessor would be a good way to get sPI times that beat conroe's :D
Click to expand...
a perfect marketing move: the XILINX bragPGA :p
 
I think it's great for AMD. I like to see them really innovating, not just producing the same (or similar) processors to Intel. If the coprocessor is successful, it will be a niche that AMD will own, even if they don't have the absolute fastest processors.
 
Jonsey said:
I think it's great for AMD. I like to see them really innovating, not just producing the same (or similar) processors to Intel. If the coprocessor is successful, it will be a niche that AMD will own, even if they don't have the absolute fastest processors.
Click to expand...

Amen to that - I am definitely looking forward to it. The only thing I don't like about it, really, is this: ;)
"Real products are at least 18 to 24 months away, if not longer, since the intention to license coherent HyperTransport was just announced today. Also, HTX slots are not part of the requirements for the first 4x4 systems, and those systems are more than likely to have both CPU sockets populated at the outset."​
- Tech Report
 
I'm with Chris in that I'm probably going Conroe within the next few months, but I am glad to see AMD using their brains to innovate rather than sit on their heels and say "we don't need a Conroe killer".

*applaud*

PS. It appears Digg has been dugg.
 
You must log in or register to reply here.
Back
Top