Benchmark Scores
The Comanche 4 benchmarking demo, software that’s about as CPU-limited as they come, was run with texture compression and hardware shaders enabled and sound disabled.
As demonstrated in our original 3.73 testing, the Prescott core’s deeper pipelines are not particularly C4 friendly, so the benchmark demo scores considerably lower than what one would expect from the clock rate of the tested CPUs. The 840 performs as expected, and with only one core effectively running the code and its lower clock speed the processor falls behind the 3.73 by roughly 10%.
Lock On: Modern Air Combat was tested using the first three minutes of the MiG-29 Intercept demo. The in-game graphics were placed at a combination of settings, with detail, scenes, and texture details at high, water at low, heat blr off, and shadows at full to increase the CPU workload.
As in the past, Lock On proves to be dependent upon the installed graphics hardware for its performance, with the two processors displaying virtually no frame rate difference. As such, the game scales rather sharply with resolution increase, losing over half its frame rate when comparing the lowest and highest tested settings. It will be interesting to see whether or not Lock On’s performance scales with next generation graphics boards.
Microsoft’s Flight Simulator 2004 was tested with high settings enabled across its four display option panels, with ground scenery cast shadows enabled. Trilinear filtering was also enabled and the max texture size slider bar placed in its middle. Sound was set to low.
The 3.73 stayed roughly 10% ahead of the 840 across the tested resolutions, which is expected considering that MS 2004 is not threaded and thus derives no benefit from the second core of the new CPU. When using our in-house demo, a fairly short dusk flight over Hong Kong city, enabling shadows cast by ground scenery drops the frame rather significantly.
Benchmark Scores
The Comanche 4 benchmarking demo, software that’s about as CPU-limited as they come, was run with texture compression and hardware shaders enabled and sound disabled.
As demonstrated in our original 3.73 testing, the Prescott core’s deeper pipelines are not particularly C4 friendly, so the benchmark demo scores considerably lower than what one would expect from the clock rate of the tested CPUs. The 840 performs as expected, and with only one core effectively running the code and its lower clock speed the processor falls behind the 3.73 by roughly 10%.
Lock On: Modern Air Combat was tested using the first three minutes of the MiG-29 Intercept demo. The in-game graphics were placed at a combination of settings, with detail, scenes, and texture details at high, water at low, heat blr off, and shadows at full to increase the CPU workload.
As in the past, Lock On proves to be dependent upon the installed graphics hardware for its performance, with the two processors displaying virtually no frame rate difference. As such, the game scales rather sharply with resolution increase, losing over half its frame rate when comparing the lowest and highest tested settings. It will be interesting to see whether or not Lock On’s performance scales with next generation graphics boards.
Microsoft’s Flight Simulator 2004 was tested with high settings enabled across its four display option panels, with ground scenery cast shadows enabled. Trilinear filtering was also enabled and the max texture size slider bar placed in its middle. Sound was set to low.
The 3.73 stayed roughly 10% ahead of the 840 across the tested resolutions, which is expected considering that MS 2004 is not threaded and thus derives no benefit from the second core of the new CPU. When using our in-house demo, a fairly short dusk flight over Hong Kong city, enabling shadows cast by ground scenery drops the frame rather significantly.