IBM published Barcelona SpecCPU results.

[hawkeyefan]

More IBM benchs (LS-DYNA benchmark) with Barcelona 1.9ghz

What's this bench?

"LS-DYNA, developed by the Livermore Software Technology Corporation, is a general-purpose transient dynamic finite element program capable of simulating complex real world problems. It is optimized for shared and distributed memory Unix®, Linux®, and Microsoft® Windows® platforms. LS-DYNA is being used by automobile, aerospace, manufacturing and bioengineering companies. LS-DYNA is heavily used in automobile crash simulation."


Result:

The x3455 server achieved an elapsed time of 31,526 seconds, which is better than the 32,578 seconds of elapsed time achieved by a single-node 8-core blade system using the Quad-Core Intel® Xeon® X5355 processor at 2.66GHz.

So, Barcelona at 1.9ghz, about 3% faster than Clovertown at 2.66ghz (with 40% less clock than Clover).

ftp://ftp.software.ibm.com/eserver/b...yna_091007.pdf

Hmm. Maybe the first decent indicator of whether Barcelona will run DC projects well compared with Xeon? That's impressive if true.

Wouldn't that mean Barcelona @ 2.4Ghz would be faster than Xeon@3.33? (on such apps, of course)

[PetNorth]

It seems it would be........

[terrace215]

Hmm. Maybe the first decent indicator of whether Barcelona will run DC projects well compared with Xeon? That's impressive if true.

LS-DYNA and their special Fluent Beta are AMD's cherry-picked benchmarks. The one niche where it looks good.

http://www.dailytech.com/article.aspx?newsid=6299

[Hornet331]

hmm that benchmarks sure seems to behave strange...

8 singlecores are faster than 2 quads....

seems, it all boils down how much nodes you use...

[PetNorth]

LS-DYNA and their special Fluent Beta are AMD's cherry-picked benchmarks. The one niche where it looks good.

http://www.dailytech.com/article.aspx?newsid=6299

yeah... superpi is much more important for human life.... who cares HPC (High Perfomance Computing), eh?

[PetNorth]

hmm that benchmarks sure seems to behave strange...

8 singlecores are faster than 2 quads....

seems, it alle boils down how much nodes you use...

I imagine it depends OS, prog version, etc, so you will have to compare similar configs, I guess...

[terrace215]

yeah... superpi is much more important for human life.... who cares HPC (High Perfomance Computing), eh?

Superpi?

How about specINT, specFP, specINT_rate? Intel wins all of these, even CLOCK for CLOCK!

Add in the clock speed advantage, the manufacturing advantage, the Q4-arriving 45nm and improved platform (Seaburg) boost, and it's hard to see how AMD digs itself out of the current hole with the K10.

[accord99]

yeah... superpi is much more important for human life.... who cares HPC (High Perfomance Computing), eh?

There are plenty of other important server benchmarks, but it clearly looks like Barcelona will only look good on HPC ones.

[PetNorth]

Superpi?

How about specINT, specFP, specINT_rate? Intel wins all of these, even CLOCK for CLOCK!

Add in the clock speed advantage, the manufacturing advantage, the Q4-arriving 45nm and improved platform (Seaburg) boost, and it's hard to see how AMD digs itself out of the current hole with the K10.

yeah yeah... or whatever

how about specFP_rate?

ah wait thats irrelevant, it is AMD cherry picked... not valid...

[JumpingJack]

yeah yeah... or whatever

how about specFP_rate?

ah wait thats irrelevant, it is AMD cherry picked... not valid...

No it is valid as a measure of floating point performance but it just is not representative of the entire pool of applications. AMD did not cherry pick, they simply choose to show their product in the best light... just as Intel shows SpecINT preferably because that is where they shine....

Focusing your attention on SpecFP and extrapolating something great across the board is simply setting yourself up for disappointment.

[JumpingJack]

LS-DYNA and their special Fluent Beta are AMD's cherry-picked benchmarks. The one niche where it looks good.

http://www.dailytech.com/article.aspx?newsid=6299

Yes.... Fluent/LS-DYNA if you notice are FP heavy... Barcey is gonna kick butt there.

[PetNorth]

I don't focuse anything. We have others to do it here.

[PetNorth]

what funny!
now rest in peace, after the effort.

[The_Beast]

Intel +1
AMD 0

[JumpingJack]

I don't focuse anything. We have others to do it here.

My bad... I thought your message implied something else.

What are your thoughts on the non-rate SpecFP and Int?

Heck, what is the difference between rate and non-rate Spec benches?

[EvlUndrWareNome]

the only thing i have heard of is Spec benches. There are other things to determine performance in apps.

Lets see a FULL REVIEW before any conclusions. Come on guys, only hours away, then the fanboys can play.

[xVeinx]

No wonder Cray and others want access to these processors.... they will eat up FPUs.

[Metroid]

September 10th where are the benchmarks, lol.

Can't wait this long.

[EvlUndrWareNome]

9:30 tonight we get to see the webcast of launch

[Metroid]

9:30 tonight we get to see the webcast of launch

London is already 10th September 01:16, hopefully it will be transmitted via internet. I'm still up to see if something comes up. So 3 hours more

It is better go to sleep and see tomorrow.

Man this is killing us.

[XeRo]

Don't they mean Sept 10 US?

[n91htmare]

Official date/time is Sept. 10, 6:45PM

So we're talking monday evening in the US.

If it's anything like the R600, which we had to wait till the end of the launch day, then I guess we'll be waiting till the end of monday

[hollo]

LS-DYNA, being a simulation program, will naturally be floating point heavy, like most simulations (eg molecular dynamics on supercomputers).

does anyone know if game physics engines use a lot of floating point? it'd seem logical, given that computer games, especially first person shooters, do a lot of simulating of real world environments & physics, and that's only becoming more common, eg ragdoll physics for dead bodies in the source engine vs icons in doom and duke nukem 3d, and lots of fun object collisions in half life 2 (gravity gun)

[dave_graham]

IBM says the results are current as of Sep 10, 2007. Sounds like they would be using the production stepping, if they bothered to put out a whitepaper on it, and why would IBM submit pre-prod stepping results to SPEC??? That just doesn't seem at all likely.

Notice the lame system availability date as well.

Barc is a paper launch from the big OEMs.

From the pdf:

Results are current as of September 10, 2007. The scores have been submitted to SPEC for review and will be posted on their Web site upon successful completion of the review. View all
published results at www.spec.org.
(1) Planned availability for the x3455 model using the AMD Opteron Model 2347 processor
(1.9GHz, 512KB L2 cache per core) is November 16, 2007.

some manufacturers have been known to toss results into the ring early...won't name names, but...this is an OEM pissing match, remember.

dave

[tsuehpsyde]

In 24 hours we'll know what's going on for sure. I just hope the Phenom products launch at higher clockspeeds (regardless of the performance). My Opteron 165 is itching to be replaced, and I'd much rather see AMD in my rig than Intel. Now if we could just get nVidia to release the 9 series...