(02-19-2013, 10:29 AM)NaturalViolence Wrote: [ -> ]Yeah that has to be why. The larger L3 cache shouldn't make a difference. Bulldozer/piledriver modules have ridiculously high L2 hit rates.
...which is why the FX-4300 with even higher clocks is considerably slower than the other Vishera CPUs?
Aren't wolfram mathematica and CS6 both multi-threaded? The FX-8350 has slightly higher clock rates as well as a slightly higher IMC speed but you're right that shouldn't be enough to account for the significant performance difference in AAC encoding. I don't know what to make of that. I guess that particular app is very memory intensive (which in retrospect makes sense). The singlethreaded benchmarks that I've seen on other sites show the FX-8350 being barely faster than the FX-4300 in singlethreaded tests. Which is what you would expect.
Thank you for the comment NaturalViolence.
NaturalViolence Wrote:Anyways I notice a few odd data points on that graph. Take a look at the 3570K results in save 1 for example.
Yeah, that IS odd. That "second" result is
mine.
Quote:Also for my results (3.7GHz) did you use my latest post where I used the GTX 660 or the older post where I used the HD 4000?
I used the results you got when you retested it with the GTX660. It's
this post.
Quote:Also I think we can safely say that something is wrong with the system of the guy you used the A10-5700 since his results are wwwaaayyy off.
Yeah, I saw his post and your comments on it. I put it in so that when he (hopefully) updates his result, it'll just be a matter of adjusting the values. If you want, I can remove it entirely.
Edit: Chart has been updated. AMD and Intel have their own colors now so it's easier to see at a glance which side does better.
Wait wait wait. I just noticed a 2500K (sandy bridge) at 4.5GHz is beating a 3570K (ivy bridge) at 4.8GHz. How the hell? Eh. One day I'll have to start doing some digging to get to the bottom of this.
When I was taking my results I waited awhile and tried to find an average fps. Usually the fps starts off higher then begins to go down a bit once the boat comes in scene 1 or 2. Maybe some people didn't wait and took the screenshot immediately.
That'll be my 2500K too.
I too waited for the average FPS without the boat in the camera view.
I think we should've had everyone take it without the boat in the camera view so we can have a fair benchmark.
Edit: Yep. Venomx1 had the boat in his camera view
when he took the screenshot. He'll definitely get a higher FPS than me when he takes one w/o the boat. I'll notify him.
So much for accuracy! Gah. We need a timedemo or something similar in the future. Doesn't dolphin support a similar feature?
I just figured any discrepancies between similar CPUs was due to one system having more background tasks running.
(02-19-2013, 03:47 PM)NaturalViolence Wrote: [ -> ]Aren't wolfram mathematica and CS6 both multi-threaded?
I DID say "lightly threaded", not single threaded for a reason.
Nevertheless, consider that even the Sandy Pentium is beating the FX-8350 in Mathematica.
Going back and looking at the pics It seems, I didn't count, pretty evenly divided on who has the boat in view and who doesn't in the first save. There is also a boat that comes into view on save 3 so the best indicator of performance is probably save 2.
Also Garteal you used my 1100t stock results from my first test which was using a badly convex stock HSF. After I put a good cooler on the cpu it brought the results up to where you would expect. ie. better than the 3.1ghz b50 I tested.
Ah, thanks for the heads up. I've updated the charts again to reflect the proper results.
Feel free to redo the benchmarks to not include the boats in the respective saves.
Don't forget to notify me so I can update the chart.
ill redo save #1. since it does seem to be inconsistant
saves #2 and 3 looks fine... except that guy that somehow outdid everybody else with his 4.4ghz 3570k. Whos is that? what Speed of ram and timings was he running. maybe an error?