Holy shit is this serious? Why are CPUs moving so fucking slow? 8350 is a budget cpu from fucking years ago
>>51609541
Moores law is slowing down as the manufacturing size decreases
>>51609541
No competition
Is a 9370 the same as a 8350 just clocked higher?
Come come a 0.4 higher clock gives 11 more fps? That's kind of alot
>>51609541
The majority of games are more GPU-intensive.
>>51609617
>>51609621
This and also a limit on how fast CPUs can get (practical limit of about 5Ghz). Improvements in architecture have helped somewhat, but the only real move forward is to add more cores and have them run at a high enough clock rate.
Unfortunately, coding for multi-threaded processes is difficult since you actually need to find a way to distribute workloads, which not all types of code are suitible for, and sometimes it's just not possible.
The few process types that do benefit from cores+threading (video/rendering) see a nearly linear improvement with cores x frequency x architecture.
Stuff like games do not utilize the CPU for more than background management of game logic, so as long as the GPU isn't being bottlenecked, the CPU has little impact on performance as long as it can throw out draw call and update the AI fast enough.
>>51609541
>Game benchmark.
There's the reason for those results.
Fucking dumbshit moron.
>>51609541
Games are for children
>>>/v/
>>51609718
Seems to have pretty huge impact if you compare the top of that chart to the bottom.
>>51609819
Bottom of the list is really the 2500K, a 2011 CPU.
4770K is 3 years newer and gains about 20fps, which isn't that much, and both are well in the realm of playability.
5960X is a $1000 CPU, it would have been disappointing if it didn't get 20~25fps more than the 4770. It can also drive both cards at 16x.
Also this test is done at 1080p with two 980 Ti cards, low resolution and SLI is an ideal test for the effect of CPU performance on games, but no one ever does that unless they're retarded.
So I need like a i7 skylake and 980 ti sli to drive a 144hz monitor at 1080p. Kind of ridiculous.
>>51609541
When properly threaded, an fx 6300 will outperform an i5 2500.
When overclocked to 4.5ghz I'm sure in battlefield 4 and crysis 3 it'll be faster than a 4460.
>6 threads
>>51609919
You need to compare the 2500K against the 4670K, and the 2600K against the 4770K mate.
Though a 2550K and 2700K would be more proper desu. So they're equal equal in frequency.
>>51610030
>that first gen Bulldozer failure
Can't believe AMD thought that'd go over well.
At least second gen was alright for workstation loads (mostly)
>>51610112
Vishera was alright.
Since the only game I play anymore is battlefield 4 I'm kinda wishing I'd saved a little money and gotten the fx 6300 instead of my i5 2400. I feel like it would have been the better choice since it would perform even better once overclocked.
