Intel Bensley Platform Preview
Posted on: 11/07/2005 06:00 AM

Test Systems and Workloads

Most of you probably know that we've been trying to expand our 2CPU benchmarking suite for a while now. While it is difficult to find workloads that scale to two threads, finding anything that scales beyond four threads is an exercise in frustration. Luckily, while we were in Oregon, Jim and I had some great opportunities to talk to some people who are experiencing the exact same issues. After much discussion and debate I've added a few workloads that I think are very pertinent, and interesting at the same time.

There have also been a lot of requests for us to add some more server-centric workloads to our reviews... We are listening! I am still tinkering with a few really interesting workloads (virtualization, database, etc) and while I had much grander plans for this article, time was working against me and I had to put a stop to "feature creep" so I would actually have time to write something for today's launch.

All that being said, let's take a look at the three boxes in the spotlight today.

I would like to take a minute to send a special thanks to Iwill (The DK88 they sent is sweet! It'll get its own review shortly.), OCZ, and PC Power & Cooling for helping me get the Opteron machine together for this test. All of the companies I just mentioned make some stellar components and I really appreciate their support. If you're looking to make some purchases for your new machine, you owe it to yourself to checkout the product line-ups from these companies.

We used the following versions of our test applications: <ul><li>SiSoft Sandra Lite 2005.SR3 (10.69)
<li>Black & Scholes Kernel (Optimized x64, custom VS2005 compile)
<li>Autodesk's 3ds Max 8.0
<li>Maxon's CINEBENCH 64-Bit Preview
<li>ScienceMark 2.0 FINAL (March 21, 2005) 64-bit
<li>Pegasys' TMPGEnc Plus v2.524.63.181
<li>Futuremark's PCMark05</ul>All benchmark numbers presented were an average of three runs, except for 3ds Max which is listed as the highest of three runs.

All test systems' desktops were set at 1024x768, 32bit color with a 75Hz refresh rate. "Optimized Defaults" were used in the BIOS on all machines. Hyperthreading was enabled on both Xeon systems for all tests.

With the exception of the Black & Scholes kernel, all of our testing methods are publicly available and easily reproducible. If you have any questions or concerns regarding the tests, feel free to discuss them in our forums.

Where are the dual-core Opterons?!?

I'm sure I will catch a ton of heat about including a pair of Opteron 246s in this comparison. The truth of the matter is... They are the only Opterons I have to test with. I have been in touch with AMD recently to plead my case and they have promised to rectify this situation very soon. I will definitely rerun all the numbers (and add a few more, I'm sure) if/when that happens.

Printed from (,2.html)