Since this is our first 8800 GTX review, I unfortunately do not have another 8800 card to compare this one with, but this will surely change in the near future as we build our 8800 GTX benchmark database with results from other branded cards. But for benchmarking sake, I will report both the stock and overclocked results. For the real-world gaming tests, I will report stock and overclocked results both at "application-controlled" image settings and 16x AF / 16xQ AA settings. For synthetic benchmarking, I will be running the 3DMark suite of programs, from 3DMark 2001 to 2006 versions as well as Aquamark3. All of the synthetic benchmarks were run at default settings. First up, 3DMark 2001.



3DMark 2001 is an aging benchmark, but it is still useful for showing how well a card's performance scales from stock to max overclocked speeds. Here we see an increase of 1,832 points. Next up, 3DMark 2003.



Again, another older benchmark but we are still able to see an increase of 3,719 points from stock to max overclock. 3DMark 2005 is next in line...



Here we gain an additional 1,143 points with our overclock. The 2006 version is our next benchmark to be run.



Our final 3DMark bench reports a gain of only 1,091 points from our baseline run. Keep in mind these scores are still pretty impressive given the overall test system specs. 3DMark 2006 Online Result Browser ranks this system's score in the top 98% of all submitted system scores / configurations. Next up we will take a run in Aquamark3. Aquamark3 was made popular in recent years and originally was a great benchmark for graphics card performance. Here lately we have seen that CPU performance greatly influences the overall score given. But for our testing that will not matter since we are not manipulating the processor clocks between tests.



Here we only gain 2,764 points over our run at stock 8800 GTX clocks. So what do all of these synthetic benchmark results tell us? Well we clearly see a gain from stock to overclocked speeds but of course that is expected. Besides being able to compare these scores to your own or other similar test configurations, they really don't give much insight into how the card would perform on a daily basis. For that, we need to obtain some real-world gaming results, and that is exactly what we will be doing...