Around June I went and purchased a Sapphire 7850 Oc Edition, I based this decision off of their High-End Video Card Benchmark Results. My final decision was ultimately going to be between the 7850 and the 560 Ti, and according to the site at the time, they were both neck and neck, but the 7850 was in the lead by just a hair.
I've frequently checked the benchmarks since then just for fun, as I like to try to stay on top of things. Unfortunately, recently I've noticed a change! The two cards are no longer neck in neck, in fact the 560 Ti is now apparently better than the 7850 AND the 7870! How is this possible? Was there some major change to the way Passmark benchmarks their games? Or maybe did NVIDIA release a massive update for the 560 Ti driver that improved its performance by a large number? What is the deal yo?
It's not reliable. If you want to check your GPU's performance, I would go on games and check it's avg and min FPS. The game that I like to check is Skyrim and BF3. Or if you want to benchmark your card solely on benchmarking software, I would go for 3DMark11 or Unigine Heaven. A 7850 would get at least P5000(Original settings on performance) and 40FPS(4xAA, 4xAF, All High, DX11, Normal Tesselation) on those tests. Which the 560 Ti gets less. It's doesn't catch up to your logic you know. Why do you have a card which is a generation before but slower? This just show how unreliable passmark tests are.
7850 is between GTX660 and 560 Ti. It's a step up from a 560 Ti for most things. Passmark is good as a general indicator, but just like any other synthetic benchmark, the results are usually not applicable to specific tasks. That's why many reviews (like the ones on Tom's) use a mix of synthetic and actual programs to benchmark.
Also, Passmark uses results submitted by end users and some people may be running extreme overclocks or unreported crossfire that can skew the results.