Sign in with
Sign up | Sign in

Benchmark Results: Sandra 2012

Intel Xeon E5-2600: Doing Damage With Two Eight-Core CPUs
By

The theoretical gains moving from two prior-generation Xeon 5600s to a pair of Xeon E5s is impressive, just as the shift from Xeon 5500 to Xeon 5600 was.

A single Core i7-3960X does extremely well compared to a pair of Xeon W5580s. However, the Xeon E5-2687Ws, based on the same Sandy Bridge architecture, benefit from an additional two cores each.

Sandra 2011’s multimedia suite similarly shows the Xeon E5s dominating. We even turned AVX instructions off to make the results more comparable. Applications optimized for the x86 extensions enjoy even greater throughput.

I didn’t bother running standalone AVX numbers this time around because the core architecture we’re dealing with here is identical to the desktop implementation. If you’d like a comparison of Intel’s AVX implementation compared to AMD’s, check out this page in AMD Bulldozer Review: FX-8150 Gets Tested, where Cakewalk’s CTO Noel Borthwick gave us access to AVX-optimized routines from Sonar X1 for testing.

Three of the CPUs in this test should support AES-NI. As I discovered when I wrote Intel Xeon 5600-Series: Can Your PC Use 24 Processors?, the company’s Xeon 5600 engineering samples didn’t yet support the feature, though. As a result, only the Core i7-3960X and Xeon E5s reflect acceleration.

Why the huge performance gap? Well, we have two processors cranking on cryptography versus one, for starters. What might you expect to see from a pair of retail Xeon 5600s in the same test? Lower performance than the E5s, almost certainly. A hardware-based feature like AES-NI is incredibly easy to execute, and we know from tests that I ran in Intel Core i7-3960X Review: Sandy Bridge-E And X79 Express that memory bandwidth is actually the bottleneck in measures of AES256 performance. Thus, a quad-channel memory controller with support for DDR3-1600 has an inherent advantage over a triple-channel controller limited to DDR3-1333.

And here’s a perfect illustration. Although registered DDR3-1600 modules are hard to come by, as mentioned on the previous page, Crucial sent over 64 GB (8 x 8 GB) of PC-12800 memory for our E5-based workstation, enabling close to two times the effective bandwidth on Xeon E5 compared to the Xeon 5600s.

Interestingly, the Core i7-3960X, armed with unbuffered DDR3-1600 is the second-place finisher, even though its four memory channels are theoretically less capable than a pair of triple-channel Xeons armed with DDR3-1333.

After back and forth emails with Adrian Silasi over SiSoftware, we couldn’t figure out why the cache performance results for the Xeon 5600-series processors were turning out so low (particularly L2 cache bandwidth, which we'd expect to be far higher). One suspicion is that this routine is tripping a throttle due to repeated use of the cache and rapidly-escalating temperatures, though Intel's engineers claim the Xeon 5500s and 5600s don't have this mechanism in place.

It’s clear, however, that the Sandy Bridge-E and Sandy Bridge-EP architectures make big improvements to L3 cache throughput by virtue of their ring buses.

Display all 80 comments.
Top Comments
  • 19 Hide
    willard , March 6, 2012 4:54 PM
    dalethepcmanNo gaming benchmarks? I know this is a high workstation / mid server build, but you know some of the boutiques will make a gaming rig out of any platform. Just out of curiosity, I would have liked to see 2x7970 or 2x580 and a few gaming benchmarks thrown in.

    I'd be really surprised to see these in gaming machines, even in the high end boutiques. That's a $2k processor they reviewed, and basically all it offers over the $1k SB-E chip (for gamers) is an extra pair of cores, which games can't make use of.
  • 18 Hide
    willard , March 6, 2012 5:24 PM
    esreverwhy aren't AMD cpus tested too? I wouldn't mind seeing how 2x interlagos stacks up.

    Anandtech benched those next to the new Xeons. Went about as well as Bulldozer vs. Sandy Bridge.

    http://www.anandtech.com/show/5553/the-xeon-e52600-dual-sandybridge-for-servers/6
  • 14 Hide
    cangelini , March 6, 2012 5:25 PM
    esreverwhy aren't AMD cpus tested too? I wouldn't mind seeing how 2x interlagos stacks up.

    Mentioned on the test page--I've invited them to send hardware and they haven't moved on it yet.
Other Comments
  • 8 Hide
    CaedenV , March 6, 2012 4:36 PM
    My brain cannot comprehend what CS5 would look like with this combined with a 1TB R4 drive, and the GTX680 version of the Quatro would look like... and I am sure my wallet cannot!

    Great article! I was not expecting my mind to be blown away today, and it was :) 
  • 19 Hide
    willard , March 6, 2012 4:54 PM
    dalethepcmanNo gaming benchmarks? I know this is a high workstation / mid server build, but you know some of the boutiques will make a gaming rig out of any platform. Just out of curiosity, I would have liked to see 2x7970 or 2x580 and a few gaming benchmarks thrown in.

    I'd be really surprised to see these in gaming machines, even in the high end boutiques. That's a $2k processor they reviewed, and basically all it offers over the $1k SB-E chip (for gamers) is an extra pair of cores, which games can't make use of.
  • 9 Hide
    nforce4max , March 6, 2012 5:07 PM
    I must say DROOL :o 

  • 14 Hide
    esrever , March 6, 2012 5:19 PM
    why aren't AMD cpus tested too? I wouldn't mind seeing how 2x interlagos stacks up.
  • 0 Hide
    reclusiveorc , March 6, 2012 5:19 PM
    I wonder how fast TempEncode would chew thru transcoding avi/wmv files to mp3/mp4
  • 18 Hide
    willard , March 6, 2012 5:24 PM
    esreverwhy aren't AMD cpus tested too? I wouldn't mind seeing how 2x interlagos stacks up.

    Anandtech benched those next to the new Xeons. Went about as well as Bulldozer vs. Sandy Bridge.

    http://www.anandtech.com/show/5553/the-xeon-e52600-dual-sandybridge-for-servers/6
  • 14 Hide
    cangelini , March 6, 2012 5:25 PM
    esreverwhy aren't AMD cpus tested too? I wouldn't mind seeing how 2x interlagos stacks up.

    Mentioned on the test page--I've invited them to send hardware and they haven't moved on it yet.
  • 10 Hide
    willard , March 6, 2012 5:32 PM
    cangeliniMentioned on the test page--I've invited them to send hardware and they haven't moved on it yet.

    I would guess that's because Interlagos is garbage compared to the new Xeons and they know it. I don't think they're terribly eager for the front page of Tom's Hardware to show the low end Xeon's beating the best Interlagos has to offer.
  • 9 Hide
    cangelini , March 6, 2012 5:47 PM
    willardI would guess that's because Interlagos is garbage compared to the new Xeons and they know it. I don't think they're terribly eager for the front page of Tom's Hardware to show the low end Xeon's beating the best Interlagos has to offer.

    Not really my place to speculate--only to point out that I similarly wanted to see AMD hardware included and explain why it isn't there :) 
  • 5 Hide
    willard , March 6, 2012 5:48 PM
    jtt283What, or who, was the target? Are there military applications for this weapon?Sorry, vote me down all you like, but the title was just silly.

    No, the title is a fairly common phrase in American English.

    "Now that I've got X, I can really do some damage" would probably be the way I hear it used most often.
  • 6 Hide
    willard , March 6, 2012 5:49 PM
    cangeliniNot really my place to speculate--only to point out that I similarly wanted to see AMD hardware included and explain why it isn't there

    Yeah, I understand that you're in a sensitive position. But being a lowly commenter, I'm free to speculate all I want!

    Muahahahaha!
  • 6 Hide
    cangelini , March 6, 2012 5:52 PM
    willardYeah, I understand that you're in a sensitive position. But being a lowly commenter, I'm free to speculate all I want!Muahahahaha!

    Precisely ;-)
  • 1 Hide
    wiyosaya , March 6, 2012 6:08 PM
    Interesting results.

    In my opinion, the SolidWorks test is also one of those not representative of typical SolidWorks tasks. PhotoView only renders realistic images of a SolidWorks model. Personally, I think the Specviewperf SolidWorks test would be significantly more representative of average SolidWorks use.

    Although I really hate to draw this comparison, PhotoView is more like using Power Point to organize a display of images created in Photoshop. In this comparison, most of the grunt work is done by Photoshop rather than Power Point, as is most of the grunt work done in SolidWorks then rendered in PhotoView. Performance differences revealed by the Specviewperf test are more informative, IMHO. See these.
  • 0 Hide
    juan83 , March 6, 2012 6:08 PM
    great review.. i wonder myself how long we 'll have to wait to see 8 cores and 16 threads on desktop segment as a default pc.. (or less than 400 dolars)

    we have to wait to long for that..
  • 0 Hide
    Anonymous , March 6, 2012 6:37 PM
    I would love one of those with a pair of FireGL cards and a mix of SCSI and SSD drives. I'm sure a dual core version of all of that will run me close to $8K though. Consider though how much Sun SPARC stations and SGI Workstations costed a decade or so ago? Workstations that were not nearly as capable went 20-25k. A dual core E5-2687 with FireGL cards and SSD drives is the fastest workstation you could put together on any platform and you can do it for far less than the 25k from years ago. Absolutely crazy to think about it in those terms.
  • 3 Hide
    EXT64 , March 6, 2012 6:52 PM
    I think you need to run some folding at home on that. I can't imagine what it would get in PPD, considering how well the old Intel 6 cores (Gulftown) do.
  • 1 Hide
    jaquith , March 6, 2012 7:11 PM
    Great article and thanks! 16-cores/32-threads is nice! :) 

    Reading this however, all I can do is think how PO'ed I am at Intel not enabling the 7th & 8th cores on the SB-E i7-3960X and i7-3930K.
  • 5 Hide
    cangelini , March 6, 2012 7:14 PM
    jaquithGreat article and thanks! 16-cores/32-threads is nice! Reading this however, all I can do is think how PO'ed I am at Intel not enabling the 7th & 8th cores on the SB-E i7-3960X and i7-3930K.

    I'm going to drop these into X79 and compare the numbers to see how power is affected. Maybe get a little overclocking out of them, just to check ;-)
Display more comments
React To This Article