Shanghai Results (Opteron 2384, 2.7GHz)

[BLaber]

Shanghai results posted by Bellismo @ xtremeforums

Just FYI:

AMD Opteron 2384: 2.7 GHz
AMD Opteron 2356: 2.3 GHz
Intel Xeon X5470: 3.33 GHz
Intel Xeon L5430: 2.66 GHz

Bench: Sungard AA
AMD Opteron 2384: 249.6s
AMD Opteron 2356: 300.2s
Intel Xeon X5470: 180.1s
Intel Xeon L5430: 230.2s

Bench: Caselab Euler 3D
AMD Opteron 2384: 63.1s
AMD Opteron 2356: 84s
Intel Xeon X5470: 64.2s
Intel Xeon L5430: 66.9

Bench: Cinebench R10
AMD Opteron 2384: 19357
AMD Opteron 2356: 15372
Intel Xeon X5470: 24624
Intel Xeon L5430: 19997

Bench: FlamMap FSPRO
AMD Opteron 2384: 427s
AMD Opteron 2356: 512s
Intel Xeon X5470: 309s
Intel Xeon L5430: 376s

Bench: Half-Life2 Build Map
AMD Opteron 2384: 91s
AMD Opteron 2356: 109s
Intel Xeon X5470: 75s
Intel Xeon L5430: 87s

Bench: HWI MySQL - Concurrency 2
AMD Opteron 2384: 1608 trans/s
AMD Opteron 2356: 1498 trans/s
Intel Xeon X5470: 1475 trans/s
Intel Xeon L5430: 1455 trans/s

Bench: HWI MySQL - Concurrency 4
AMD Opteron 2384: 3072 trans/s
AMD Opteron 2356: 2948 trans/s
Intel Xeon X5470: 2753 trans/s
Intel Xeon L5430: 2741 trans/s

Bench: HWI MySQL - Concurrency 8
AMD Opteron 2384: 5646 trans/s
AMD Opteron 2356: 4944 trans/s
Intel Xeon X5470 4633 trans/s
Intel Xeon L5430: 4648 trans/s

Bench: HWI MySQL - Concurrency 16
AMD Opteron 2384: 5961 trans/s
AMD Opteron 2356: 4286 trans/s
Intel Xeon X5470 5846 trans/s
Intel Xeon L5430: 4766 trans/s

Bench: HWI MySQL - Concurrency 32
AMD Opteron 2384: 6123 trans/s
AMD Opteron 2356: 4970 trans/s
Intel Xeon X5470: 6046 trans/s
Intel Xeon L5430: 4838 trans/s

Bench: PovRay 3.7b21 - Chess 1280*1024
AMD Opteron 2384: 58.4s
AMD Opteron 2356: 69.4
Intel Xeon X5470 48.1s
Intel Xeon L5430: 59.1s

Bench: Sissoft Sandra 2009 - Intercore Bandwidth
AMD Opteron 2384: 7.46 Gbit/s
AMD Opteron 2356: 5.79 Gbit/s
Intel Xeon X5470: 25.7 Gbit/s
Intel Xeon L5430: 21.9 Gbit/s

Bench: Sissoft Sandra 2009 - Intercore Latency
AMD Opteron 2384: 107ns
AMD Opteron 2356: 164ns
Intel Xeon X5470 104ns
Intel Xeon L5430: 101ns

Bench: Sissoft Sandra 2009 - L1 latency
AMD Opteron 2384: 3clk
AMD Opteron 2356: 3clk
Intel Xeon X5470 3clk
Intel Xeon L5430: 3clk

Bench: Sissoft Sandra 2009 - L2 Latency
AMD Opteron 2384: 16clk
AMD Opteron 2356: 16clk
Intel Xeon X5470: 18clk
Intel Xeon L5430: 18clk

Bench: Sissoft Sandra 2009 - Memory Bandwidth
AMD Opteron 2384: 20.41 Gbit/s
AMD Opteron 2356: 17.75 Gbit/s
Intel Xeon X5470: 4.05 Gbit/s
Intel Xeon L5430: 4.05 Gbit/s

Bench: Sissoft Sandra 2009 - Memory Latency
AMD Opteron 2384: 105 ns
AMD Opteron 2356: 118 ns
Intel Xeon X5470: 126 ns
Intel Xeon L5430: 127 ns

Idle Power Concumption
AMD Opteron 2384: 100W
AMD Opteron 2356: 109W
Intel Xeon X5470: 170W
Intel Xeon L5430: 169W

Power Consumption (Cinenbench)
AMD Opteron 2384: 197W
AMD Opteron 2356: 226W
Intel Xeon X5470: 277W
Intel Xeon L5430: 227W

Power Consumption (HWI SQL CC 32)
AMD Opteron 2384: 143W
AMD Opteron 2356: 167W
Intel Xeon X5470: 206W
Intel Xeon L5430: 188W

Enjoy

 

[Gikaseixas]

Not bad at all if true

 

[MarcVenice]

Well, although Intel is still outperforming AMD's offerings in roughly half the benchmarks ( didn't really count ), because of the lower powerconsumption, it could be a real winner for server use ...

 

[BLaber]

Tomorrow is the rumored release date for shanghai , so well get to know soon...

Shanghai can now be purchased : http://www.it4profit.com/ecata...ND=701&SDATA=1&LANG=en

OS2376WAL4DGIWOF CPU Server Opteron Quad Core Model 2376 (6MB,75W,Socket F) box 429.00
OS2378WAL4DGIWOF CPU Server Opteron Quad Core Model 2378 (6MB,75W,Socket F) box $595.00
OS2380WAL4DGIWOF CPU Server Opteron Quad Core Model 2380 (6MB,75W,Socket F) box $795.00
OS2382WAL4DGIWOF CPU Server Opteron Quad Core Model 2382 (6MB,75W,Socket F) box $994.00
OS2384WAL4DGIWOF CPU Server Opteron Quad Core Model 2384 (6MB,75W,Socket F) box $1,126.00
OS8384WAL4DGIWOF CPU Server Opteron Quad Core Model 8384 (6MB,75W,Socket F) box $2,445.00
OS8384WAL4DGI CPU Server Opteron Quad Core Model 8384 (6MB,75W,Socket F) tray $2,446.00

 

[BLaber]

Originally posted by: MarcVenice
Well, although Intel is still outperforming AMD's offerings in roughly half the benchmarks ( didn't really count ), because of the lower powerconsumption, it could be a real winner for server use ...

With the expected Higher clocked Bins in 2009 Q1 , performance gap will come down, also then will come HT3.1 , DDR3 :thumbsup:

Above test are done using HT1.0

 

[Sylvanas]

Actually, AMD win the majority of benchs posted above 13-7- looks good, hope Deneb comes through.

 

[SlowSpyder]

Originally posted by: Sylvanas
Actually, AMD win the majority of benchs posted above 13-7- looks good, hope Deneb comes through.

It looks pretty competitive, but a lot of those 'wins' are just really measurements (like L2 latency) or are synthetic. Hopefully this will translate into real world performance for AMD, but at any rate it does look like a good improvement over the prvious 2MB L3 cache model.

 

[Philippart]

you forgot to include this info, which makes the results even better:

Just FYI:

AMD Opteron 2384: 2.7 GHz
AMD Opteron 2356: 2.3 GHz
Intel Xeon X5470: 3.33 GHz
Intel Xeon L5430: 2.66 GHz

For clock/clock comparisons: Optie 2384 vs Xeon L5430

 

[BLaber]

Shanghai demolishes 6 core intel duninghton :Q :thumbsup:

 

[SlowSpyder]

Originally posted by: BLaber
ftp://ftp.software.ibm.com/ese...S42_specjbb_111308.pdf">Shanghai demolishes 6 core intel duninghton</a> :Q :thumbsup:

ftp://ftp.software.ibm.com/ese...S42_specjbb_111308.pdf

I think that's the link.

 

[Lonyo]

What RAM will Nehalem for servers use? Since one of the serious negative points for current Xeons is the use of FB-DIMM's in terms of power consumption, will Nehalem do anything to swing performance/watt back in Intel's favour?

 

[BLaber]

Originally posted by: SlowSpyder

Originally posted by: BLaber
ftp://ftp.software.ibm.com/ese...S42_specjbb_111308.pdf">Shanghai demolishes 6 core intel duninghton</a> :Q :thumbsup:

<a target=_blank class=ftalternatingbarlinklarge href="ftp://ftp.software.ibm.com/eserver/benchmarks/news/newsblurb_LS42_specjbb_111308.pdf">ftp://ftp.software.ibm.com/......specjbb_111308.pdf</a>

I think that's the link.

OOPS my bad didn't check the link after posting it , thz for correct working link SlowSpyder

 

[JackyP]

Originally posted by: Lonyo
What RAM will Nehalem for servers use? Since one of the serious negative points for current Xeons is the use of FB-DIMM's in terms of power consumption, will Nehalem do anything to swing performance/watt back in Intel's favour?

I've heard nehalem will sport ECC DDR3, so yes this should help.

 

[nyker96]

I wonder how the shanghai core achieve such low power consumption under load/idle, that's almost unbelievable numbers in the power department. but clearly the chip is optimized for server load like SQL, as for workstation loads like cinebench or home loads like gaming. It's still a tad behind intel's offerings.

 

[Accord99]

Originally posted by: nyker96
I wonder how the shanghai core achieve such low power consumption under load/idle, that's almost unbelievable numbers in the power department. but clearly the chip is optimized for server load like SQL, as for workstation loads like cinebench or home loads like gaming. It's still a tad behind intel's offerings.

Mainly the chipset and using lower power consuming R-DDR2 instead of FB-DIMMs. I'd still expect Yorkfield to have lower power consumption versus Deneb.

 

[Zstream]

Wait what?!

How could I have been so right when others stated otherwise. All along I stated that the chip will perform well in the server market with the faster and larger l3 cache.

Oh snap!

Oh and to the FB dimm mention, on a normal FB dimm configuration you do not consume 65+ watts...

 

[Cookie Monster]

How does one measure intercore bandwidth?? AMD seems to suffer alot when it comes to this synthetic benchmark. Any ideas on why?

Shanghai is shaping up nicely. Along with HT3.0 and other refinements to the process (faster clocked shanghai's hopefully), this chip is going to be very competitive in the server market.

 

[Idontcare]

The power consumption at load is tasty

 

[classy]

Looking at the similar results and the similar clock speeds, there appears to be very little difference. If those numbers are true AMD looks to be able to challenge Intel straight up. I remember a web page not to long ago that said Deneb should be a great gaming cpu. Looking at these numbers if true, looks like it may be all good for AMD.

 

[Accord99]

Originally posted by: Zstream
Oh and to the FB dimm mention, on a normal FB dimm configuration you do not consume 65+ watts...

Well clearly for this configuration, the chipset and/or FB dimms do consume a significant amount of power, otherwise a Barcelona system would not be power competitive with a system that uses low-voltage Harpertowns that consume less than 50W per CPU.

 

[nyker96]

Originally posted by: Accord99

Originally posted by: nyker96
I wonder how the shanghai core achieve such low power consumption under load/idle, that's almost unbelievable numbers in the power department. but clearly the chip is optimized for server load like SQL, as for workstation loads like cinebench or home loads like gaming. It's still a tad behind intel's offerings.

Mainly the chipset and using lower power consuming R-DDR2 instead of FB-DIMMs. I'd still expect Yorkfield to have lower power consumption versus Deneb.

so that means on the desktop, this power advantage will be non-existent since both company uses DDR2 or DDR3s?

 

[Accord99]

Originally posted by: nyker96
so that means on the desktop, this power advantage will be non-existent since both company uses DDR2 or DDR3s?

Yes, I personally expect Yorkfield to have noticeable power advantage at load versus a comparable performing Deneb. While it looks like Deneb will show a great reduction in power usage compared to the current Phenoms, Intel's 45nm CPUs also had a great reduction in power consumption. For example here:

http://www.behardware.com/arti...ore-2-q9300-e7200.html

 

[Nemesis 1]

Well its all good . Heres some projected numbers from IC7 . We don't want Harptown setting there all alone.

http://www.xtremesystems.org/f...howthread.php?t=178416

 

[mrSHEiK124]

Originally posted by: Cookie Monster
How does one measure intercore bandwidth?? AMD seems to suffer alot when it comes to this synthetic benchmark. Any ideas on why?

Shanghai is shaping up nicely. Along with HT3.0 and other refinements to the process (faster clocked shanghai's hopefully), this chip is going to be very competitive in the server market.

I think it's because all the launch day Shanghai stuff is only running HT1.0. Isn't my Socket 939 Athlon 64 X2 of yester-generation running at HT2.0? That would explain a whole lot.

 

[nyker96]

Originally posted by: Accord99

Originally posted by: nyker96
so that means on the desktop, this power advantage will be non-existent since both company uses DDR2 or DDR3s?

Yes, I personally expect Yorkfield to have noticeable power advantage at load versus a comparable performing Deneb. While it looks like Deneb will show a great reduction in power usage compared to the current Phenoms, Intel's 45nm CPUs also had a great reduction in power consumption. For example here:

http://www.behardware.com/arti...ore-2-q9300-e7200.html

In any case Anand has a new article on Shanghai up as well. from look of things there's no denial it's a monster of a chip for server market. the question for me personally is how the consumer versions of the chip perform. that we will probably know next year. at least the 45nm process seems alright for amd.