Linked by Thom Holwerda on Fri 4th Dec 2009 19:40 UTC
Post a Comment
Let's Hope AMD manage to take at least 50% market share with Fusion. Looking possible with Nvidia with no product, prototypes being slower than AMD/ATI's released product and Intel making little progress on the CPUGPU front.
If that doesn't happen? Well it's gonna be pretty bad for us consumers. Intel has even pushed Nvidia out of making chipsets for it now.
A platform controlled by a single company; especially one that has had to copy x86-64 and on die memory controllers from its competitor just to progress, is bad for everyone
Edited 2009-12-05 14:56 UTC
Wow, dishonest much?
Given that there has been not a single performance number released for Fermi, and yet you can make claims regarding its performance vs ATi's current line.
Given that there has been not a single performance number released for Fermi, and yet you can make claims regarding its performance vs ATi's current line.
The Specs released by nvidia and its latest clock speeds on their A2 Silicon put it at a theoretical peak of 1TFlop maximum performance. The current released single ATI card reaches 2.7TFlops - this is a completely different ballpark. So no this is based on reality and actual data. Not your assumptions based on nothing. I suggest you read the latest Fermi Whitepaper.
And that you hope that a non-existant AMD product (fusion) will be able to crush the CPU-GPU single package with working silicon by intel...
Intel's solution has not even reached GT200 performance levels for a product set to be released in a years time. It was also recently cancelled because it "just wasnt good enough". Intel has never had a good graphics solution, it still doesn't. There is only one company that has both a viable CPU and GPU design. Guess who?
Edit: I should probably also add that Fusion was working in 2009 but and I quote "wasn't compelling enough to release" on 45nm.
Edited 2009-12-07 23:32 UTC



