Intel Quad-core: Clovertown Performance Reviewed
Posted on: 11/11/2006 06:00 AM

Test Systems

Since everyone's focus is turning towards power consumption, and "performance-per-watt" these days, I made a special effort to level the playing field as much as I could this time around. A single test system was used for all three sets of Intel processors. Nothing at all changed between platforms except for the actual CPUs, but I'll get into that in more detail later in this article.

The Opteron platform I used is the same one that I've presented in previous articles. It is based on an Iwill DK88 motherboard, OCZ memory, a pair of Opteron 285 (2.6ghz dual-core) processors and a PC Power & Cooling PSU. The Opteron 285 is currently one of the fastest dual-core AMD processors available, but the (now) dated Socket 940 platform might not perform quite as well as a similar Socket F configuration. Unfortunately, my attempts to acquire numbers from a suitable Socket F platform fell through at the last minute.



In all of the following graphs, the platforms will be called by the official processor product names (Opteron 285, Xeon 5160 and Xeon E5345). In the case of the Dempsey Xeons, my CPUs are preproduction and run at a frequency of 3.46ghz, and there are no production parts that run at this speed. I will reffer to these preproduction Dempseys throughout this article as "Xeon 50xx" as the Dempsey line of processors are also known as the "Xeon 5000 Series".

Workloads

I used the following versions of our test applications: <ul><li>SiSoft Sandra 2007 SP1
<li>Black & Scholes Kernel (Optimized x64, custom VS2005 compile)
<li>Autodesk's 3ds Max 8.0
<li>SunGard Adaptiv Credit Risk Analysis 3.0
<li>Pegasys' TMPGEnc Plus v2.524.63.181
<li>POV-Ray 3.70 beta 16a
<li>SPECjbb2005
<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 the "Xeon 50xx" machine 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


Printed from 2CPU.com (http://www.2cpu.com/contentteller.php?ct=articles&action=pages&page=intel_quad_core_clovertown_performance_reviewed,2.html)