I want to wait now (small, screaming, smelly thing may get in the way of PC build project in a couple of weeks:D) until January/February before making my final decision on a new build. I am currently looking at the Q6600, but could be swayed if the 45nm processors (particularly the quad core parts) are just as cheap and work just as well.
Of course, the wild card is AMD. If they bring out a killer chip at a good price point, I may just remain an AMD fanboy
Any thoughts on what's coming from AMD?
Andy
clock for clock k10 quad cores are about 5% behind intels 45nm quads in crysis according to the number floating around the web... well, we dont know prices yet... so who has the better price / performance ratio remains to be seen
Nope. Tim just spoke to Intel on the phone and launch dates get decided literally a couple of weeks or days before actual launch. It's the same with everyone, that's why you get ambiguous "Q" numbers.
Quote: "I guess you could think of the QX9650 as the Bugatti Veyron of the CPU world - it annihilates everything in its path."
What about the IBM POWER6? Or the Sun UltraSPARC T2? I don't think that I'd go so far to say that it annilhilates EVERYTHING in its path.
Amazing that even with the AMD chips using a process that's twice as big, it only represents about a 25% reduction in power consumption (on the new Intel chips). I would have expected it to be more. It'd be interesting (just as a point of reference) to see how the power consumption numbers stack up against an Opteron 2350.
P.S. Your charts are mislabeled (starting pg. 15) where the QX9650 is listed as 2X 3.0 GHz. FYI.
Definitely a very fast processor, I'll grant you that.
Originally Posted by alpha754293 Quote: "I guess you could think of the QX9650 as the Bugatti Veyron of the CPU world - it annihilates everything in its path."
What about the IBM POWER6? Or the Sun UltraSPARC T2? I don't think that I'd go so far to say that it annilhilates EVERYTHING in its path.
Amazing that even with the AMD chips using a process that's twice as big, it only represents about a 25% reduction in power consumption (on the new Intel chips). I would have expected it to be more. It'd be interesting (just as a point of reference) to see how the power consumption numbers stack up against an Opteron 2350.
P.S. Your charts are mislabeled (starting pg. 15) where the QX9650 is listed as 2X 3.0 GHz. FYI.
Definitely a very fast processor, I'll grant you that.
Charts - fixed.
Lets not be pedantic about what processors - it's a desktop CPU and currently the fastest on the market. It's not Xeon, although it can be "workstation". There are technicalities to every argument ;) If we really wanted to nit pick we could look at GPGPU throughput and core FLOPS processing, software optimisations or not even x86 stuff.
Originally Posted by alpha754293 Amazing that even with the AMD chips using a process that's twice as big, it only represents about a 25% reduction in power consumption (on the new Intel chips). I would have expected it to be more. It'd be interesting (just as a point of reference) to see how the power consumption numbers stack up against an Opteron 2350.
Surely it is impressive regarding the power consumption, seeing as the intel chip is a quad, and the 6400 et al are dual cores? Unless I'm missing something here...?
Originally Posted by Bindibadgi Well overclocking is entirely CPU (and motherboard) dependent and these are ES chips remember, not retail. We may simply just be lucky (for a change).
Heli - we could spend all day/all year doing every different bit of software but we have covered video encoding with two different codecs and software, as well as image manipulation using Paint.NET :)
Roto - it's just representative mate, don't read into it too much. EDIT: Ahh crap!! I completely forgot I've got a ton of die pics on the camera so I'll whip them out and we can see what the actual dies look like.
Afaik Intel is supposed to release several 45nm processors in Jan, then later (as yields increase and depending on what the competition does) more speed versions. Where'd you read March?
I was pleased with the review Bindibadgi . . . it was the performance increase of the latest 45nm processor that left me a little dissapointed. I appreciate the video and image editing tests you did, and appreciate that with so many different possible tests no one review can cover all variables . . . but what I meant to say in general was that a photoshop shop benchmark relevent to my use was the test I am most looking forward to seeing (from whatever source that may be).
Comments 26 to 39 of 39
clock for clock k10 quad cores are about 5% behind intels 45nm quads in crysis according to the number floating around the web... well, we dont know prices yet... so who has the better price / performance ratio remains to be seen
What about the IBM POWER6? Or the Sun UltraSPARC T2? I don't think that I'd go so far to say that it annilhilates EVERYTHING in its path.
Amazing that even with the AMD chips using a process that's twice as big, it only represents about a 25% reduction in power consumption (on the new Intel chips). I would have expected it to be more. It'd be interesting (just as a point of reference) to see how the power consumption numbers stack up against an Opteron 2350.
P.S. Your charts are mislabeled (starting pg. 15) where the QX9650 is listed as 2X 3.0 GHz. FYI.
Definitely a very fast processor, I'll grant you that.
Charts - fixed.
Lets not be pedantic about what processors - it's a desktop CPU and currently the fastest on the market. It's not Xeon, although it can be "workstation". There are technicalities to every argument ;) If we really wanted to nit pick we could look at GPGPU throughput and core FLOPS processing, software optimisations or not even x86 stuff.
Surely it is impressive regarding the power consumption, seeing as the intel chip is a quad, and the 6400 et al are dual cores? Unless I'm missing something here...?
Remember a 90 SOI vCore is still only 1.45ish volts, whereas a 45nm chip is 1.1/2ish. You don't get a halving of power reduction.
Rich, is this in your offices or a libary piccey?
I was pleased with the review Bindibadgi . . . it was the performance increase of the latest 45nm processor that left me a little dissapointed. I appreciate the video and image editing tests you did, and appreciate that with so many different possible tests no one review can cover all variables . . . but what I meant to say in general was that a photoshop shop benchmark relevent to my use was the test I am most looking forward to seeing (from whatever source that may be).
Thanks for the article.
It is only a refresh though ;) hopefully when the mid-range retail versions arrive they'll overclock by a load so there will be extra difference :D
(I see why you asked now, the table had an error in it - should be fixed now!)