- May 19, 2011
- 20,395
- 15,085
- 136
Example:
http://anandtech.com/bench/product/538?vs=647
The way that I read the benchmark results is that for all the actual gaming benchmarks, the 660 Ti manages double or triple the FPS that the 5770 can do. In one ray test (pov-ray I assume this is), the 5770 wins. In the power consumption and noise level tests, the results are one way or the other, usually pretty close.
Methinks the overall summary graph at the top is being incorrectly thrown by the power/noise results, which surely should carry less of a weight on the overall result than the rest. Is anyone wanting to upgrade from a 5770 for gaming performance reasons really going to say "well, the 660 Ti is hardly any better because it's a bit more noisy and uses a bit more power under load"? Doesn't that statement pretty much apply to every newer graphics card since the 5770 (or at least as a general rule)?
Of course a summarised result from a set of benchmarks isn't going to hold 'absolute truth', but I think the summary in this case has little or no relationship to reality. It also doesn't seem to be the conclusion of the standard reviews.
I do appreciate the effort that goes into the 'bench' part of the Anand site, I find it extremely useful in some situations instead of having to mix and match the results from various articles to compare say a two-year-old product with a new one from not necessarily the same marketing bracket.
http://anandtech.com/bench/product/538?vs=647
The way that I read the benchmark results is that for all the actual gaming benchmarks, the 660 Ti manages double or triple the FPS that the 5770 can do. In one ray test (pov-ray I assume this is), the 5770 wins. In the power consumption and noise level tests, the results are one way or the other, usually pretty close.
Methinks the overall summary graph at the top is being incorrectly thrown by the power/noise results, which surely should carry less of a weight on the overall result than the rest. Is anyone wanting to upgrade from a 5770 for gaming performance reasons really going to say "well, the 660 Ti is hardly any better because it's a bit more noisy and uses a bit more power under load"? Doesn't that statement pretty much apply to every newer graphics card since the 5770 (or at least as a general rule)?
Of course a summarised result from a set of benchmarks isn't going to hold 'absolute truth', but I think the summary in this case has little or no relationship to reality. It also doesn't seem to be the conclusion of the standard reviews.
I do appreciate the effort that goes into the 'bench' part of the Anand site, I find it extremely useful in some situations instead of having to mix and match the results from various articles to compare say a two-year-old product with a new one from not necessarily the same marketing bracket.
Facebook
Twitter