Page 1:Making Every Dollar Count
Page 2:CPU And Cooler
Page 3:Motherboard And Memory
Page 4:Graphics Card and Hard Drive
Page 5:Case, Power Supply, And Optical Drive
Page 7:Test System Configuration And Benchmarks
Page 8:Benchmark Results: Crysis, Unreal Tournament 3
Page 9:Benchmark Results: World In Conflict, Supreme Commander
Page 10:Benchmark Results: Audio/Video Encoding
Page 11:Benchmark Results: Applications
Page 12:Benchmark Results: Synthetics
As mentioned, the $500 PC was built around the idea of overclocking and squeezing as much performance possible out of each dollar spent. Before we see just what speeds this system is capable of reaching, let’s look at the stock settings.
Our GA-EP35-DS3L motherboard shipped with the F4 BIOS and no matter how high VCore was set in the BIOS, we were unable to get any voltage setting to take above 1.296 V. Limited to the normal VCore, the E2180 topped out at just 2.66 GHz (10*266), which wasn’t going to cut it for our System Builder Marathon. The solution was to flash the BIOS to version F5, which then made it possible to bump up the VCore and see just what this chip could do.
At 1.504 V, our E2180 was passing any synthetic or gaming tests attempted at 3.33 GHz (10*333), but was unable to pass our stress testing. Rather than bump the voltage up beyond 1.5 V, we lowered our goals. Dropping the CPU clock ratio (multiplier) to eight and raising the host frequency to 400 MHz, we reached a stable 3.2 GHz at 1.456 V and a +0.1 V bump in the FSB voltage. At these settings, memory is kept at 800 MHz (1:1) and no attempts were made to add more voltage to the Wintec RAM and tighten the rather relaxed SPD timings.
Having found our maximum stable CPU speed, we turned our attention to getting more performance from the PNY 8800GT. Even with its single-slot cooler, we found stability at 741 MHz core and 1,890 MHz shaders. That is quite a boost from the factory speeds of a 600 MHz core and 1,512 MHz shaders. For testing, we backed both down one notch to a 738 MHz core and 1,836 MHz shaders.
On the memory side, the sky was the limit and we had no problems at our highest tested speed of 1,065 MHz (2,130 MHz effective). Choosing a speed to run the memory at for our testing presented a problem, since although we had not reached a limit on the RAM chips themselves, many people feel that the 8800 GT itself can’t handle these memory speeds for long periods of time and it’s safer for the card to keep the gaming speeds of the memory well below 2.0 GHz. We still wanted to squeeze as much performance as we felt comfortable running, so we knocked the memory down a few notches and decided on 1,053 MHz (2,106 MHz effective) for testing.
It’s worth noting that the extra boost in memory data rate speeds from 1,980 MHz to 2,106 MHz provided less than 300 extra 3DMarks and just a fraction of one frame per second (FPS) in our Crysis testing. Not all games may respond the same as Crysis did, but with such minimal gains seen here, it indeed doesn’t appear to be worth pushing the 8800 GT’s RAM to these speeds for daily gaming.
- Making Every Dollar Count
- CPU And Cooler
- Motherboard And Memory
- Graphics Card and Hard Drive
- Case, Power Supply, And Optical Drive
- Test System Configuration And Benchmarks
- Benchmark Results: Crysis, Unreal Tournament 3
- Benchmark Results: World In Conflict, Supreme Commander
- Benchmark Results: Audio/Video Encoding
- Benchmark Results: Applications
- Benchmark Results: Synthetics