Westmere-EX: Intel's Flagship Benchmarked
by Johan De Gelas on May 19, 2011 1:30 PM EST- Posted in
- IT Computing
- Intel
- Xeon
- Cloud Computing
- Westmere-EX
vApus Mark II
vApus Mark II is our newest benchmark suite that tests how well servers cope with virtualizing "heavy duty applications". We explained the benchmark methodology here. We used vSphere 4.1 Update 1, based upon the 64 bit ESX 4.1.0 b348481 hypervisor.
* with 128GB of RAM
Benchmarks cannot be interpreted easily, and virtualization adds another layer of complexity. As always, we need to explain quite a few details and nuances.
First of all, we tested most servers with 64GB of RAM. However, the memory subsystem of the Quad Xeon needs 32 DIMMs before it can deliver maximum bandwidth. As some of these server systems will get those 32 DIMMs while others will not, we tested both with 16 (64GB) and 32 DIMMs (128GB). Our vApus mark test requires only 11GB per tile: 4GB for the OLTP database, 4GB for the OLAP and 1GB for each of the three web applications (3GB in total). So even a five tile test demands only 55GB. Thus, in this particular benchmark there is no real advantage to having 128GB of RAM other than the bandwidth advantage for the quad Xeon platform. That is why we do not test the the Quad Opteron with more than 64GB: it makes no difference and makes the graph even more complex.
Then there's the problem that every virtualization benchmark encounters: the number of tiles (a tile is a group of VMs). With VMmark, the benchmark folks add tiles until the total throughput begins to decline. The problem with this approach is that you favor throughput over response time. In the real world, response time is more important than throughput. We test with both four (20 VMs, 72 vCPUs) and five tiles (25 VMs, 90 vCPUs). Which benchmark gives you the most accurate number for a given system? Let us delve a little deeper and take the response time into account.
Facebook
Twitter
RSS
62 Comments
View All Comments
Fallen Kell - Thursday, May 19, 2011 - link
As the subject says. Would love to see how these deal with something like Linpack or similar.erple2 - Thursday, May 19, 2011 - link
I'd be more interested at seeing how they perform in slightly more "generic" and non-GPU optimizeable workloads. If I'm running Linpack or other FPU operations, particularly those that parallelize exceptionally well, I'd rather invest time and money into developing algorithms that run on a GPU than a fast CPU. The returns for that work are generally astounding.Now, that's not to say that General Purpose problems work well on a GPU (and I understand that). However, I'm not sure that measuring the "speed" of a single processor (or even a massively parallelized load) would tell you much, other than "it's pretty fast, but if you can massively parallelize a computational workload, figure out how to do it on a commodity GPU, and blow through it at orders of magnitude faster than any CPU can do it".
However, I can't see running any virtualization work on a GPU anytime soon!
stephenbrooks - Thursday, May 19, 2011 - link
Yeah, well, in an ideal world...But sometimes (actually, every single time in my experience) the "expensive software" that's been bought to run on these servers lacks a GPU option. I'm thinking of electromagnetic or finite element analysis code.
Finite element engines are the sort of thing that companies make a lot of money selling. They are complicated. The commercial ones probably have >10 programmer-years of work in them, and even if they weren't fiercely-protected closed source, porting and re-optimising for a GPU would be additional years work requiring programmers again at a high level and with a lot of mathematical expertise.
(There might be some decent open-source alternatives around, but they lack the front ends and GUI that most engineers are comfortable using.)
If you think fixing the above issues are "easy", go ahead. You'll make millions.
L. - Thursday, May 19, 2011 - link
lolif you code .. i don't want to read your code
carnachion - Friday, May 20, 2011 - link
I agree with you. In my experience GPU computing for scientific applications are still in it's infancy, and in some cases the performance gains are not so high.There's still a big performance penalty by using double precision for the calculations. In my lab we are porting some programs to GPU, we started using a matrix multiplication library that uses GPU in a GTX590. Using one of the 590's GPU it was 2x faster than a Phenon X6 1100T, and using both GPUs it was 3.5x faster. So not that huge gain, using a Magny-Cours processor we could reach the performance of a single GPU, but of course at a higher price.
Usually scientific applications can use hundreds of cores, and they are tunned to get a good scaling. But I don't know how GPU calculations scales with the number of GPUs, from 1 to 2 GPUs we got this 75% boost, but how it will perform using inter-node communication, even with a Infiniband connection I don't know if there'll be a bottleneck for real world applications. So that's why people still invest in thousands of cores computers, GPU still need a lot of work to be a real competitor.
DanNeely - Saturday, May 21, 2011 - link
single vs double precision isn't the only limiting factor for GPU computing. The amount of data you can have in cache per thread is far smaller than on a traditional CPU. If your working set is too big to fit into the tiny amount of cache available performance is going to nose dive. This is farther aggravated by the fact that GPU memory systems are heavily optimized for streaming access and that random IO (like cache misses) suffers in performance.The result is that some applications which can be written to fit the GPU model very well will see enormous performance increases vs CPU equivalents. Others will get essentially nothing.
Einstein @ Home's gravitational wave search app is an example of the latter. The calculations are inherently very random in memory access (to the extent that it benefits by about 10% from triple channel memory on intel quads; Intel's said that for quads there shouldn't be any real world app benefit from the 3rd channel). A few years ago when they launched cuda, nVidia worked with several large projects on the BOINC platform to try and port their apps to CUDA. The E@H cuda app ended up no faster than the CPU app and didn't scale at all with more cuda cores since all they did was to increase the number of threads stalled on memory IO.
Marburg U - Thursday, May 19, 2011 - link
Finally something juicy,JarredWalton - Thursday, May 19, 2011 - link
So, just curious: is this spam (but no links to a separate site), or some commentary that didn't really say anything? All I've got is this, "On the nature of things":http://en.wikipedia.org/wiki/De_rerum_natura
Maybe I missed what he's getting at, or maybe he's just saying "Westmere-EX rocks!"
bobbozzo - Monday, May 23, 2011 - link
Jarred, my guess is that it is spam, and that there was a link or some HTML posted which was filtered out by the comments system.Bob
lol123 - Thursday, May 19, 2011 - link
Why is there a 2 socket only line of E7 (E7-28xx), but at least as far as I can tell, not any 2-socket motherboards or servers? Are those simply not available yet?