• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Official Thread - Nehalem performance estimates

I

Idontcare

Elite Member
http://blogs.zdnet.com/Ou/?p=1025

Sort of an extrapolated normalized comparison of Intel leaked slides on Nehalem performance versus Intel's estimates of Shanghai's performance.

Is Nehalem really going to deliver this kind of performance? I am 100% skeptical given the past history of how these SpecFP and SpecINT numbers get played out before product releases...

http://www.anandtech.com/trade...howdoc.aspx?i=2956&p=5

edit: per newschool's request, thread title changed to "official"
 
I tend to read George Ou with a grain of salt, as I've found him to be somewhat one-sided with his reporting of AMD/Intel. Calling him the Intel version of Tom Yager might be a little too much, but I still read a bit of Intel bias in his writing. That's not to say he hasn't been right before, of course.
 
Yeah he is a big fan of Intel, to put it mildly. But these estimates are from Intel. Personally I think it looks resonable considering the improvements in Nehalem over previous gen Penryn derivates. Keep in mind that these are SPEC int/fp RATE for 2-socket systems (Xeon DP). "rate" scores mean they run multiple copies of the same benchmark at the same time and mesure throughput. SMT and the massive bandwidth improvements of Nehalem should ensure a massive boost in this particular test. AMD has always done well in SPECfp rate for 2+ sockets thanks to the integrated memory controller on each CPU in the system.
 
So Nehalem has been shown running a full OS, while Shanghai has yet to. Is this correct?

And with that he says it is therefore unlikely for Shanghai to release at the same time as Nehalem. Both scheduled for Q4-08/Q1-09 are they not?
 
Originally posted by: AmberClad
I tend to read George Ou with a grain of salt, as I've found him to be somewhat one-sided with his reporting of AMD/Intel. Calling him the Intel version of Tom Yager might be a little too much, but I still read a bit of Intel bias in his writing. That's not to say he hasn't been right before, of course.
Click to expand...

just a bit of intel bias?
 
SPECfp is very dependant on bandwith throughput. Considering Nehalem will have tripple channel DDR3 on the QPI bus, I'd say these results are plausible.

Either way, it would be quite embarassing for Intel if they don't meet these targets once Nehalem ships. 😉
 
Originally posted by: KingstonU
So Nehalem has been shown running a full OS, while Shanghai has yet to. Is this correct?

And with that he says it is therefore unlikely for Shanghai to release at the same time as Nehalem. Both scheduled for Q4-08/Q1-09 are they not?
Click to expand...

Shanghai is scheduled for a ramp in H1 of 08, and volume ship for revenue in H2 of 08. Shipments are supposed to begin in mid 08...

"News from last week?s AMD conference call that told us the company would start production of 45nm microprocessors in the first half of 2008 shouldn?t be looked at from the point of view of ?volume production.? The message that should be taken from the news is that at some time in the first six months of next year, AMD may well have a small amount of devices entering the supply-chain.

AMD had said that we should see 45nm chips ?mid-year,? which could easily have applied to the time period from June through September based on Barcelona schedules"



 
I've not been paying attention for while now, got sucked into consoles. And I right in thinking:

Nehalem is scheduled for about October 2008 and it's probably going to carry the early-adopter price premium until what March 2008? And also paying a premium on the mobo and ram. But, buying a Quad now means both the mobo and ram is useless on the next upgrade, right?
 
Originally posted by: Davegod
I've not been paying attention for while now, got sucked into consoles. And I right in thinking:

Nehalem is scheduled for about October 2008 and it's probably going to carry the early-adopter price premium until what March 2008? And also paying a premium on the mobo and ram. But, buying a Quad now means both the mobo and ram is useless on the next upgrade, right?
Click to expand...

I think you meant March 2009, but if Shanghai doesn't given Yorkfield a run for the performance crown then I doubt very much Intel will pursue desktop Nehalem with any sort of aggressiveness.

Being a monolithic die, Nehalem will have the same sort of Achilles heel that Phenom and Itanium have in regards to die-size versus yield. If Intel has an option to continue shipping MCM quad-cores for significant revenue in 1H/09 then I'd expect shareholders to be irritated if they don't.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top