Currently I am benchmarking my 4770K and FX-8350. I have owned the 4770K since it release (within the first month of its release in the UK) and I owned the 8350 a while back (release day) but at that time I didnt have it for very long, I was, erm lets say I know a lot more now. Over the years I have noticed benchmarkers altering or being super biased, not to metion any names but they can be easy to spot at time. Even youtube benchmarks that I didnt even expect that everyone seems to like (im not talking about logan/teksydicate lol just in case people started thinking, if they were I would be here). But at that time I didnt really care to be honest but it did bug me at times, but most of all the things that bugged me were super fan boys, for example NV fanboys still try to defend the Titan Z, yes the 3K graphic card.
One day someone was selling their 8350, it was lapped and at that time I needed a second PC. I decided to buy it again and see how it performs, if its sucks just play around with the chip (I like to overclock at times) I got the chip and noticed my frames were really similar with BF4 on both chip (at that time I was playing BF4). Will add more soon...
If you want to just see the benchmarks the game you only play then go down below. If your game isnt there just comment with the game and Ill add it to my list ;)
If you get the opportunity next week watchdogs would be a cool one to bench. Though I dont want to push you to go out and buy a new game just to bench.
Good to see by the way, that there is allmost no diffrence between both cpu´s in all those games, BF4 on mantle, even does better, on amd
Even in Metro Last light, which is a cpu demending game, the diffrence is only 5 fps, with both cpu´s on stock speeds. This is the most interessting benchmark.
you could also bench skyrim, this is also discussed alot. probably an intel winner, but would like to know by how much.
Also interessting is the Watchdogs benchmark, this game is stated as a multithreaded game, and thats exactly what we see, the 4770K and 8350 perform roughly the same. And that makes sense.
ARMA and DayZ are just crap gaming engine´s, they run badly on amd but also badly on an intel i5/i7 it will perform the same as on an i3. not realy games to botter about in my opinnion.
I have ARMA 3 somewhere on my drive, but i dont realy botter about that game.
The older Metro 2033 performs better on FX 8350 than any ivy bridge intels on Teksyndicate's benches using a hd 7870.
I played through it on high but on dx10 on a 4870 1gb which would be a hd 7750 equivalent.
If the stakes were up this wouldn't be a fair comparison as FX 8350 is a bit older than haswell and while the difference is small an FX 8350 would have to be overclock to match a 4670 @ stock.
I wish AMD would make a proper FX proc like they used to
@Fatal, should the standard deviation be measured? Suppose for example, the Intel average is 72 frames per second and the AMD average is 68. But the Intel frame rate dips below 60 sometimes whereas AMD does not. Which chip gives higher performance in that case?
If there's a widely known method to interpret average fps near 60, I haven't seen it spelled out.
I can imagine a person buying the chip that gives the higher average frame rate, and having a worse game experience as a result.