Sign in with
Sign up | Sign in
Your question

Intel Shows off 45nm Penryn V8

Last response: in CPUs
Share
June 6, 2007 5:19:04 PM

I don't think this has been posted yet, so here it is:




http://www.legitreviews.com/article/521/1/
June 6, 2007 5:26:00 PM

Wow. That was something I wasn't expecting to see.

I still think V8 was knee-jerking reactionary to something that really didn't need to be addressed (4X4), but if they can make the V8 run without the use of FB-DIMMS, then it could be a viable workstation replacement (I am still doubtful about desktop replacement).
June 6, 2007 5:38:32 PM

Anyone know if Cinebench 10 uses SSE4?
Related resources
June 6, 2007 5:43:36 PM

A nice +10% performance per clock over the Q6800.
June 6, 2007 5:55:40 PM

Quote:
Wow. That was something I wasn't expecting to see.

I still think V8 was knee-jerking reactionary to something that really didn't need to be addressed (4X4), but if they can make the V8 run without the use of FB-DIMMS, then it could be a viable workstation replacement (I am still doubtful about desktop replacement).


Well to be picky. It's something thats not needed, but it is something that needed to be addressed from Intel's position. Can't let AMD have any market, even if it is a stupid market with no point 8)
June 6, 2007 6:04:21 PM

"AMD wouldn't comment on the record on what their upcoming Barcelona processor would score, but people in the know said that AMD is getting ~16,000 with a pair of their Barcelona processors on Cinebench 10. If this proves to be true and you keep in mind that the 45nm Penryn V8 system is underclocked right now and scoring ~23,000 on the same benchmark it means things could get interesting really soon."

Nasty ...
June 6, 2007 6:10:44 PM

double post
June 6, 2007 6:14:22 PM

Quote:
"AMD wouldn't comment on the record on what their upcoming Barcelona processor would score, but people in the know said that AMD is getting ~16,000 with a pair of their Barcelona processors on Cinebench 10. If this proves to be true and you keep in mind that the 45nm Penryn V8 system is underclocked right now and scoring ~23,000 on the same benchmark it means things could get interesting really soon."

Nasty ...



Well, clockspeed maybe the only thing that Penryn has over Barcelona as the Anand test shows that Barcelona was running at half the clock and others are saying that the fastest right now is at 2GHz (1 GHz slower).

Had they mentioned what clocks that 16000 came from it would give a better picture. But from that statement, AMD is behind by ~ 31% and clockspeed can be estimated at 33% slower ( at 2GHz - even more at 1.6GHz) so it looks like it will be a two horse race.
June 6, 2007 6:38:10 PM

Quote:
"AMD wouldn't comment on the record on what their upcoming Barcelona processor would score, but people in the know said that AMD is getting ~16,000 with a pair of their Barcelona processors on Cinebench 10. If this proves to be true and you keep in mind that the 45nm Penryn V8 system is underclocked right now and scoring ~23,000 on the same benchmark it means things could get interesting really soon."

Nasty ...



Well, clockspeed maybe the only thing that Penryn has over Barcelona as the Anand test shows that Barcelona was running at half the clock and others are saying that the fastest right now is at 2GHz (1 GHz slower).

Had they mentioned what clocks that 16000 came from it would give a better picture. But from that statement, AMD is behind by ~ 31% and clockspeed can be estimated at 33% slower ( at 2GHz - even more at 1.6GHz) so it looks like it will be a two horse race.
Yeah, but the reason AMD is demoing at 1.6GHz is that they can't get the clock speed much higher and still get past POST.

Quote:
One vendor demonstrated Barcelona to DailyTech running at 1.6GHz. Current AMD Barcelona samples are not scaling too well. AMD partners confirmed the highest running, POST and OS capable, Barcelona processor is 2.0 GHz. AMD previously posted benchmarks of a simulated 2.6 GHz Barcelona.

So, while I don't disagree with your statement that "clockspeed maybe the only thing that Penryn has over Barcelona", it could be a big big problem for Barcelona if they don't start getting the clockspeed up a bit. Just my 2 cents. :) 
June 6, 2007 7:20:40 PM

Quote:
So, while I don't disagree with your statement that "clockspeed maybe the only thing that Penryn has over Barcelona", it could be a big big problem for Barcelona if they don't start getting the clockspeed up a bit. Just my 2 cents.



That's what I said. I don't mean to take anything from Penryn, but you see what I mean about the numbers?

AMD says they will release up to 2.3GHz ( I thought it was 2.5GHz) and if the IMC is causing the problem then the new respin could do as much as the one for AM2 and C2D last year.

Plus AMD uses the CTI method (Intel seems to be adopting it also) so they may hit 3GHz by 2008. A real bright spot is that Shanghai and Montreal will be mostly just shrinks to 45nm so it should be much smoother than Barcelona as it's a redesigned architecture.
June 6, 2007 9:51:29 PM

Fiar enough.

Anyway, yeah if Barcy hits 3GHz I imagine it would be a pretty sweet deal. My guess is that a lot of how well it does will have to do with availability and the actual street price. Anyway, I guess we'll just have to wait and see what happens.
June 6, 2007 11:50:36 PM

Nice :D 

Barcelona getting owned by Penryn is pretty sad, since Nehalem is going to be the main competition. Another sign that Intel is not slowing down. From the looks of things now that we are well into 2007, AMD as we know is most likely over with. Not being able to demo at at least a 2.5 GHz part at the 65 nm node speaks volume about the immaturity of Barcelona design
June 7, 2007 1:02:26 AM

Quote:
Anyone know if Cinebench 10 uses SSE4?


http://www.lostcircuits.com/cpu/intel_penryn2/

24.9% advantage for Penryn in 9.5 and 25.5% in 10... You can also see that quad Penryn scales against dual Penryn better in Penryn. Funny though, in 10, Kentsfield has bigger advantage against dual Penryn :) 
June 7, 2007 1:44:24 AM

i like the case they used, really sweet 8)
June 7, 2007 1:51:28 AM

barcey over all seems like a rush job. and i hope that does not hinder its overall popularity.
June 7, 2007 2:16:22 AM

There’s no doubt that penryn will be hard to compete with but then again we really don’t know what’s up behind closed doors at AMD.
I like a good race but its no fun if you already know who the winner is.
Hopefully AMD will surprise us.
June 7, 2007 2:25:11 AM

Quote:
There’s no doubt that penryn will be hard to compete with but then again we really don’t know what’s up behind closed doors at AMD.
I like a good race but its no fun if you already know who the winner is.
Hopefully AMD will surprise us.


thats exactly what people were saying about intel when netburst flopped. and then out of no where they unleashed a new core that is putting heat back on AMD.

I am convinced AMD has alot more to offer than what they are showing at this point. very convinced.
June 7, 2007 2:29:24 AM

I certainly hope so but then again that Penryn will fit nicely in my P180 8)
June 7, 2007 3:07:33 AM

If Barcelona doesn't improve its efficiency per clock a lot we will have no competition. Even if Barcelona hits 3Ghz looks like it won't be enough to beat Penryn, maybe just enough to be in the same league.

Remember that are rumors that it could clock Penryn to close 4 GHz and yorkfield to 3.5 GHz. It's a lot. And we should consider Nehalem, the real competitor to Barcelona. We can't guess how good it will be, but better than penryn/yorkfield it's for sure.
June 7, 2007 3:40:32 AM

Quote:
I am convinced AMD has alot more to offer than what they are showing at this point. very convinced.

Why? What makes you convinced?
June 7, 2007 4:50:14 AM

Quote:
thats exactly what people were saying about intel when netburst flopped. and then out of no where they unleashed a new core that is putting heat back on AMD.

I am convinced AMD has alot more to offer than what they are showing at this point. very convinced.


Uh, the Core microarchitecture didn't come out of nowhere.

Even a year before launch, Intel was already yelling out codenames like Conroe and Woodcrest. At the same time, they gave out hints like 4-issue wide execution, memory disambiguation, and shared L2. They pointed at Dothan and Yonah and highlighted performance/watt.
!