Originally posted by: Sonikku
The article paints a very inaccurate picture.
The problem with "accuracy" and benchmark comparisons is that you can often change 1 setting with any given game and produce results that look vastly different from what another website has posted.
It's always been this way, and it will always be this way.
What you need to do is look at a wide variety of sites, and study how each came up with their respective results. For example, in that Anandtech article, you have to go all the way to the back to see the settings they used for the games-- not one used AA. Some were run at 1024x768. They did this in order to not put most of the stress on the GPU, and allow CPU scaling to show the differences between the chips. That makes sense in a CPU article, right?
If you find yourself playing STALKER at 1024x768 without AA, then these results might be useful for you. For the rest of us, we are forced to look elsewhere for something more comparable to what we do.
For example, let's use
Firing Squad. They compare the same 2 chips, and when you look at the results for HL2 Ep1, you'll see that at lower resolutions without AA, that the e6750 is a stronger chip and performs better. But at 1600x1200 with AA, the difference is 1 or 2%. The AT article showed it much higher (again because of the lack of AA).
Take *all* benchmarks with a grain of salt (or a whole bottle if you're browsing HardOCP). Any specific benchmark setup and run-through are likely not indicative of everyone's real-world scenario, as components vary wildly and 2 setups are rarely identical. Look for websites that show a wide variety of games being tested under several resolutions and settings, with different CPUs and video cards.
The price difference doesn't make the comparison "unfair", since prices fluctuate over time. In fact, the comparison in general isn't "unfair". We should compare processors of different speeds and prices. It's important to know where one excels and another falls short. It's important to see that even a low-end X2 3800+ can compete very well with a C2D at 3.9ghz (if gaming is your focus). I don't just want to know how my $150 processor compares to other $150 chips... I want to know how it compares to $1000 chips and $70 chips too. I just want to know if I'm getting the best value for my money. Only very thorough benchmarks can ensure me of that. And, as you can see, you won't find that from just a single website.
Facebook
Twitter