It will be intersting see more details about this processor. What is the target product for such processor, might be a server. AMD is also seems to be positive with their fusion series processors made to compete with Intel Atom.
To @goafrit: I don't think that there is any danger of Intel forced to break up. Yes, AMD might not be doing that great but they will not disappear. After spinning Global Foundries (which is doing great on its own) they will probably do somewhat better going forward. And the true competitor emerged: ARM. Finally, their Microsoft alliance and related monopoly has been broken. All in all Intel has to fight hard to continue to be the dominant player...dr Kris
You had better hope and pray that Intel doesn't put AMD out of business because if they do then the FTC and Justice Department will break Intel into little tiny pieces for being a monopoly.
The last time Intel was the only provider the CPUs and Match Coprocessors to go with them were way overpriced but Intel had the whole show. Now that AMD is producing x86 CPUs compatible with Intel and there is pressure on both sides to keep the prices down it is good for everybody.
One more thing, why is Intel even still putting resources into the Itanic? If you go back to the "way back" machine the Itanium was Intels sole answer to 64 bit computing.
It was only after AMD gained a foot hold selling its mixed 32/64 bit CPUS (that Intel said they would never do) did Intel start producing 32/64 bit CPUS.
While AMD had the 32/64 bit CPU domain to itself it seriously started eating into Intels bottom line.
So you can thank the competition with AMD for the Core Duo and the follow-up i7 series which replicated AMD in using a "HyperTransport" type serial bus instead of the old bottleneck "Intel Hub" architecture.
So the next time you purchase ANY CPU, thank AMD for keeping Intels prices down.
thank you David but I remain a little skeptical here...it remains rather well known that performance doesn't scale linearly with number of processors, the goal of wholly grail of parallel processing has not be achieved...sure,you can pack 8 cores but how many are working in parallel all the time...Kris
"The quad core wasn't even close to linear speed increase"
what do you mean, name these apps/code bases that dont perform "to linear speed increase" please
for instance x264 performed 'almost' perfectly linear speed increases across the board, but then they dont totally rely on the compiler to provide the improved speed as x264 dev's write and test every single SIMD optimisation for a given C routine OC
sure you may try and say/imply that this is the exception to the rule , but clearly only relying on the compiler to provide the improved speed is a really bad idea, rather than have your dev's always write and test SIMD and even simple style changes to sections of code to improve the throughput in any market segment including server's
simply asking the question is the compiler producing actual sane and faster assembly output, (you did actually check it right ?), is a good thing to always do , where are those bottlenecks, hunt them down and remove them , and then say with confidence you dont get a linear speed increase at least up to 12 or 16 cores.
Eight processors? At the end of the day what kind of performance improvement will you have? The quad core wasn't even close to linear speed increase, I'll be more excited to see development platforms that can actually exploit multiple processors without the nightmare debugging.
My Mom the Radio Star Max MaxfieldPost a comment I've said it before and I'll say it again -- it's a funny old world when you come to think about it. Last Friday lunchtime, for example, I received an email from Tim Levell, the editor for ...
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...