Testing Methodology
At Icrontic, we do our best to ensure that we are consistent and thorough in our testing. Below is an outline of our testing methodology and configuration.
Comparison to Other Processors
Finding processors to compare the Phenom X3 to was definitely a challenge. Since there are no triple core processors on the market, comparing the X3 to dual and quad core processors was our only option. We didn’t have a lot of time to prepare for this review, so we had to work with what was available to us. None the less, it should provide a good performance comparison across the varying product lines. We’ll be trying to compare processors in similar price ranges.
Based on the pricing we have for the Phenom X3s, we’ll be using two Intel Core2 Duo processors—the E8400 and the E8200. The E8400 is priced at about $205 today—assuming you can find it anywhere. This is about $5-10 more than our X3 8750. Not a perfect comparison, but pretty close in price. The E8200 is priced at around $180-190. The X3 8650 should retail somewhere in the $175-$180 range, just a little below the E8200. We really don’t have anything to compare the $150 X3 8550 to, so it will be interesting to see how it fairs along side its big brothers. The soon to be released Intel E7200 may be a worthy adversary, but we do not have one in the lab at this time.
Since we only have the X3 8750 in the lab, we’ll be simulating the other two models by reducing the CPU multiplier to 11.5 and 10.5 for the 8650 and 8450 respectively. The performance provided in these simulated configurations will be identical–only power consumption may differ, which is why we only use the default configurations for those tests. A similar simulation will be done to produce the E8200 by reducing the multiplier of the E8400–again, it will perform identically to its retail counterpart.
We’ll be throwing AMD X2 tests into the mix as well to see how Phenom X3 compares to its dual-core predecessors.
System Configuration
AM2 Platform:
- AMD Phenom X3 8750 (2.4GHz) or AMD Athlon 64 X2 6000+ (3.0GHz)
- Noctua NH-U12P Heatsink with 1600RPM Scythe S-FLEX fan.
- Gigabyte GA-MA78GM-S2H Motherboard (780G Chipset)
- 2x1024MB Corsair XMS2 PC2-8500 DDR2 modules
- Seagate 7200.7 Hard Drive
- Pioneer DVR212D Optical Drive
- Corsair TX750 Power Supply
775 Platform:
- Intel E8400 Dual-Core (3.0GHz, 6MB L2 cache) or Intel Q6600 (2.4GHz, 2x4MB L2 cache)
- Noctua NH-U12P Heatsink with 1600RPM Scythe S-FLEX fan.
- Asus P5K-E Motherboard (Intel P35 chipset)
- 2x1024MB Corsair XMS2 PC2-8500 DDR2 modules
- Seagate 7200.7 Hard Drive
- Pioneer DVR212D Optical Drive
- Corsair TX750 Power Supply
The system memory is run at 1066MHz on both the Phenom X3 and Intel Core2 processors. Timings are manually set to 5-5-5-18-2T. The X2 processors can not run 1066MHz natively, so the memory is set to 800MHz, with the same timings. Due to the nature of the IMC on the X2 processors, the memory frequency runs slightly lower than 800MHz (750MHz with our 6000+). After conducting some preliminary testing, leaving the Phenom memory controllers ‘unganged’ for dual 64bit operation seemed to yield the best results in a variety of benchmarks.
For CPU testing moving forward, we’ve decided to use Windows Vista. Vista Ultimate is installed and patched with SP1. The following performance oriented changes were made:
- System Restore Disabled
- Windows Sidebar Disabled
- Windows Defender Disabled
- Superfetch Disabled*
- Aeroglass Theme Disabled
- UAC Disabled
*Although Superfetch is designed to improve performance, it can add inconsistency to results that is difficult to control. Rather than trying to ‘train’ Superfetch and having to run benchmarks numerous times to get the best result, I disabled the Superfetch service.
Additional details surrounding the specific tests done will be outlined in their respective sections. Ambient temperature is maintained at 19°C for all power consumption tests.