Im impressed with the amd fx 8 core cpu

[gasolin]

Im impressed with the amd fx 8 core cpu, i have a fx 8300 at 4.0ghz on all cores, 16gb 1866mhz ram and the usually ssd, 

mine is the very popular samsung 840 evo 

 

Atleast 30 tabs, 2x twitch, 1 cam chat with 2 open cams (not my cam) and youtube playing videos, all with chrome

and i do not have more than 50-65% cpu usage with 1x 2560x1440 and 1x 3440x1440 monitor + a palit gtx 1060 6gb 1620mhz

 

My cpu score in passmark at 4.0 ghz is about 9000 or just above 9000 and i know a haswell cpu at 3.7ghz on all cores only score about 8100

 

I wouldn't mind having a ryzen setup if it where to cost as littel as the fx 6/8 core cpus (dare i ask what the rumor is about the ryzen prices compared to the fx &/8 core cpus?)

 

Oh forgot yesterday i watched tech talk with JayzTwoCents and Barnacules Nerdgasm and toook cinebench benchmark and all was smooth as butter and a week or 2 ago i play a bit project cars it ran as it should, when i was finished i noticed i had a minimized twicth live stream running in the background,didn't noticed any of that playing all smooth

[refinedskillz]

The 8 cores really does kick in during multitasking, but gaming performance on standalone isnt really that good.

[DarkEnergy]

It's not that great of a CPU. It's alright for multitasking but not for games. 

[meenmeen1103]

Yeah, the fx 8 series is pretty good at multitasking, but gets bashed a lot for lower ipc and less fps than intel varieties. Until you want 1440p 144hz you'll probably not care to upgrade.

[dizmo]

I can do the same thing on my i5 - 3317 with 4GB of RAM

Hurray for dual cores.

[79wjd]

1 minute ago, dizmo said:

I can do the same thing on my i5 - 3317 with 4GB of RAM

Hurray for dual cores.

+1

 

Thats not a very heavy load.

[meenmeen1103]

Just now, djdwosk97 said:

+1

 

Thats not a very heavy load.

You're crazy, a single chrome tab maxes out 4gb ram...

[CantThinkOfAUserName]

6 minutes ago, meenmeen1103 said:

Yeah, the fx 8 series is pretty good at multitasking, but gets bashed a lot for lower ipc and less fps than intel varieties. Until you want 1440p 144hz you'll probably not care to upgrade.

Depends if the game even uses hyperthreading/more then 4 cores. If not its not going to help you for 144hz gaming. Resolution doesnt really affect cpu usage though.

[79wjd]

1 minute ago, CantThinkOfAUserName said:

Depends if the game even uses hyperthreading/more then 4 cores. If not its not going to help you for 144hz gaming. Resolution doesnt really affect cpu usage though.

FX cpus have trouble pushing 144fps in a lot of games where even an i5 has no trouble. And resolution definitely affects cpu usage as the load switches to the GPU as the resolution increases. Thats why CPU gaming benchmarks are usually done at resolutions like 800x600.

[meenmeen1103]

1 minute ago, CantThinkOfAUserName said:

Depends if the game even uses hyperthreading/more then 4 cores. If not its not going to help you for 144hz gaming. Resolution doesnt really affect cpu usage though.

A game not utilizing all cores, plus the relative lower gaming performance of the fx due to ipc&freq will lower fps, but also leaves room for streaming programs to run smoothly. The 1440p part will probably be heavy for a 1060 on dual monitors though.

[gasolin]

Im just surprised that in passmark i get around 9000 and my old i5 4590s that was limited to 3.7ghz on all cores got almost 1000 points lower

 

In geeekbench 4 the intel i5 crushes the fx 8 core i total score, where intel always gets better single core performance no matter what benchmark i take

 

Aida cpu queen they are about the same 3.7 and 4.0 ghz  mabye a slightly advantage for the fx 8 core at 4.2ghz

 

I had to lower a few setting (gtx 1060 6gb) in need for speed shift 2 unleashed than my usual intel setting to get in the late 80's or higher fps, but project cars ran without having to change one settings

 

Goes to show that newer games can handle more cores,threads better than an old game, that's what i felt, who would have thought since i need for speed shift 2 unleashed is a much older game than projectcars and you can't compare it to the demanding older games like crysis 1-3, that a newer game ran better than a old game in the same genre

[gasolin]

21 minutes ago, dizmo said:

I can do the same thing on my i5 - 3317 with 4GB of RAM

Hurray for dual cores.

Do you have 2 high resolution monitors? 2560x1440 and 3440x1440 

[dizmo]

8 minutes ago, gasolin said:

Do you have 2 high resolution monitors? 2560x1440 and 3440x1440 

That doesn't really affect Windows.

And, to be fair, I don't even have a dedicated GPU. I'm running on HD4000 integrated graphics.

[Glenwing]

3 hours ago, djdwosk97 said:

FX cpus have trouble pushing 144fps in a lot of games where even an i5 has no trouble. And resolution definitely affects cpu usage as the load switches to the GPU as the resolution increases. Thats why CPU gaming benchmarks are usually done at resolutions like 800x600.

Resolution doesn't affect CPU load. CPU benchmarks are done at low resolutions specifically because it doesn't affect CPU load, so it doesn't interfere with the results of how many frames the CPU can generate.

 

The load doesn't "switch" to the GPU at higher resolutions. The CPU load required per frame remains the same, while the GPU load per frame increases. If you are GPU bottlenecked at a higher resolution, then yes the total load on the CPU "decreases", but that's only because the framerate has been lowered by the GPU. If you had infinite GPU power, this would never be the case, and the framerate would be limited to the CPU's maximum limits, which would be exactly the same at any resolution (given the same aspect ratio). So if your CPU limits you to 80 fps at 720p, it will also be limited to 80 fps at 8K. Of course if your GPU is not strong enough, then that will impose an even lower limit on top of that (maybe 30 fps), and now the CPU load will "decrease" because it isn't required to draw as many frames, but the load per frame is the same.

 

This situation is the definitive "GPU bottleneck". If you were to upgrade your GPU, a more powerful GPU might raise that 30 fps limit to 50 fps, or 70 fps. But if you get a GPU ten times more powerful than that, you will only get up to 80 fps and stop, because now you have the "CPU bottleneck" situation. If you have enough GPU power to achieve the CPU's maximum framerate, that limit will be the same at any resolution.

[CantThinkOfAUserName]

8 minutes ago, Glenwing said:

Resolution doesn't affect CPU load. CPU benchmarks are done at low resolutions specifically because it doesn't affect CPU load, so it doesn't interfere with the results of how many frames the CPU can generate.

 

The load doesn't "switch" to the GPU at higher resolutions. The CPU load required per frame remains the same, while the GPU load per frame increases. If you are GPU bottlenecked at a higher resolution, then yes the total load on the CPU "decreases", but that's only because the framerate has been lowered by the GPU. If you had infinite GPU power, this would never be the case, and the framerate would be limited to the CPU's maximum limits, which would be exactly the same at any resolution (given the same aspect ratio). So if your CPU limits you to 80 fps at 720p, it will also be limited to 80 fps at 8K. Of course if your GPU is not strong enough, then that will impose an even lower limit on top of that (maybe 30 fps), and now the CPU load will "decrease" because it isn't required to draw as many frames, but the load per frame is the same.

 

This situation is the definitive "GPU bottleneck". If you were to upgrade your GPU, a more powerful GPU might raise that 30 fps limit to 50 fps, or 70 fps. But if you get a GPU ten times more powerful than that, you will only get up to 80 fps, because now you have the "CPU bottleneck" situation. If you have enough GPU power to achieve the CPU's maximum framerate, that limit will be the same at any resolution.

Thank you for this information you said it a lot better than I could.

[Mr.Meerkat]

Ah, it's time for my favourite 2 benchmarks that tilts Intel lovers off earth  

http://www.ocaholic.co.uk/modules/smartsection/item.php?itemid=4003&page=7

Before someone quotes me again saying how the 6600K is faster in the majority of benchmarks, let's see, a 4~ year older and almost half as expensive is beating the newer and more expensive chips in a couple of benchmarks, I wonder...

 

Seriously, FX chips are not bad for their price at all, they're great streaming and workstation chips where it's competitor is an i3 6100/i5 6400...(price wise).

[79wjd]

34 minutes ago, Glenwing said:

-snip-

Resolution does still affect the CPU usage though. It just does it through a GPU bottleneck (or lack thereof). Technically it's the GPU that causes the CPU load to be affected and not the resolution, but the resolution affects the GPU load. 

[gasolin]

Im surprised that there isn't a bigger difference in this video, since at stock speed the i7 4790k has 2/3 more single core performance or 1000 points than a fx 8350 Passmark

 

A r9 390 is actually as fast as a gtx 980/1060 3gb 

 

[Glenwing]

37 minutes ago, djdwosk97 said:

Resolution does still affect the CPU usage though. It just does it through a GPU bottleneck (or lack thereof). Technically it's the GPU that causes the CPU load to be affected and not the resolution, but the resolution affects the GPU load. 

Yes, but the point is the resolution has no effect on how much CPU power is required to achieve a certain framerate. I know what you actually said was "resolution affects CPU usage" which is technically true, but I just think it's important to word things carefully, because there's a lot of belief starting to spread around that "higher resolutions take less CPU power", because people misinterpret the statement "CPU is less important at higher resolutions due to GPU bottlenecks". And when you said the load "switches" from the CPU to the GPU you seemed to be implying the first.