Ryzen 1700 and 1080 bad performance

[wiseoxejah]

I'm coming from this thread because I have determined that it is not the gpu. I decided to ask about the entire system in a new thread. If you read through that thread you'll see what I've tried so far. I will try anything and will pay for a working answer.

 

[photonboy]

I'm coming from this thread because I have determined that it is not the gpu. I decided to ask about the entire system in a new thread. If you read through that thread you'll see what I've tried so far. I will try anything and will pay for a working answer.

Not sure why you made a new thread as I followed the link, however you should CLOSE that one and stick with one thread.

Here's my GTX1080 actual results for a benchmark. you must use the EXACT SAME SETTINGS listed:
Unigine Valley
-DX11, Ultra, x8AA, full screen, 2560x1440

GTX1080
- 68.4FPS avg
- 2863
- 32.7fps low
- 140.3fps high

If your results are close then your GPU is fine.

your comments suggest there may be a CPU or System memory bottleneck.

1) DDR4 - should be at least 3000MHz Dual Channel to avoid much bottleneck in some situations (if so, should be fine)

2) CPU - monitor the CPU frequency while stressing with something like Handbrake. You can try Prime95 too as long as you don't overheat. You should be hopefully getting at least 3.7GHz or so depending on if you overclock or not.

Other:
Games at lower resolutions (i.e. 1080p) can be noticeably lower FPS scores than same GTX1080 with something like OC'd i7-7700K CPU.

So.. if the GPU works, and the CPU frequency can be maintained under load etc then I'd say it's about whether the game benefits from a faster CPU or not (at that resolution/settings).

 

[photonboy]

You can also PM me if you want, however a response may not be timely.

No need to offer to pay. I'm not sure if that's even appropriate here.

So we need to know EXACTLY why you think your performance is lower in a repeatable situation. Preferably an offline game/benchmark to avoid networking as part of the equation.

 

[rgd1101]

Please stick with 1 thread.

 

[wiseoxejah]

Please stick with 1 thread.

Apologies for that. I was just getting desperate and annoyed.

 

[wiseoxejah]

You can also PM me if you want, however a response may not be timely.

No need to offer to pay. I'm not sure if that's even appropriate here.

So we need to know EXACTLY why you think your performance is lower in a repeatable situation. Preferably an offline game/benchmark to avoid networking as part of the equation.

I just ran userbenchmark and it appears to be my cpu. Results are here. I will run any game or benchmark you'd like. I'll do the Heaven one now. UPDATE: PerformanceTest says my ram is running at 1501.5mhz. Idk what to do about that. It says in my task manager that is is 2933.

 

[wiseoxejah]

I'm coming from this thread because I have determined that it is not the gpu. I decided to ask about the entire system in a new thread. If you read through that thread you'll see what I've tried so far. I will try anything and will pay for a working answer.

Not sure why you made a new thread as I followed the link, however you should CLOSE that one and stick with one thread.

Here's my GTX1080 actual results for a benchmark. you must use the EXACT SAME SETTINGS listed:
Unigine Valley
-DX11, Ultra, x8AA, full screen, 2560x1440

GTX1080
- 68.4FPS avg
- 2863
- 32.7fps low
- 140.3fps high

If your results are close then your GPU is fine.

your comments suggest there may be a CPU or System memory bottleneck.

1) DDR4 - should be at least 3000MHz Dual Channel to avoid much bottleneck in some situations (if so, should be fine)

2) CPU - monitor the CPU frequency while stressing with something like Handbrake. You can try Prime95 too as long as you don't overheat. You should be hopefully getting at least 3.7GHz or so depending on if you overclock or not.

Other:
Games at lower resolutions (i.e. 1080p) can be noticeably lower FPS scores than same GTX1080 with something like OC'd i7-7700K CPU.

So.. if the GPU works, and the CPU frequency can be maintained under load etc then I'd say it's about whether the game benefits from a faster CPU or not (at that resolution/settings).

My results:
-43.0FPS avg
- 1800
- 9.0fps low
- 89.6fps high

 

[photonboy]

Something is really wrong with those results. The CPU should not be a bottleneck, so assuming you had the EXACT same settings your GPU must be getting stuck at a lower frequency.

Possibly at around 1200 to 1300MHz?

Try running the benchmark again whilst monitoring with MSI Afterburner, EVGA Precision OC or suitable tool that has an OSD to overlay BOTH of these:

1) GPU frequency, and
2) GPU usage

Normally in that benchmark you should see about 95% usage and 1900MHz+ frequency on the GPU (except between scene transitions).

You might have for example, 1250MHz frequency and high GPU usage. (You should get the same thing if the CPU was the bottleneck but it would have to be really bad to be a bottleneck for this benchmark.)

*If you manually overclocked, try doing this:
a) remove any software used to OC or monitor the graphics card
b) run DDU http://www.guru3d.com/files-details/display-driver-uninstaller-download.html
c) reinstall the video drivers
d) rerun the benchmark

if still wrong, then
e) verify CPU frequency under load with Task Manager ("CPU Speed" should say about 3.7GHz?) for say Prime95 (just enable half the threads)

(right-click START)-> TM-> Performance-> CPU-> "Speed" (it should vary a lot in idle usage, then be pretty stable under load)

f) if still wrong (I mean if CPU results are good), contact the graphics card manufacturer

 

[photonboy]

You can also PM me if you want, however a response may not be timely.

No need to offer to pay. I'm not sure if that's even appropriate here.

So we need to know EXACTLY why you think your performance is lower in a repeatable situation. Preferably an offline game/benchmark to avoid networking as part of the equation.

I just ran userbenchmark and it appears to be my cpu. Results are here. I will run any game or benchmark you'd like. I'll do the Heaven one now. UPDATE: PerformanceTest says my ram is running at 1501.5mhz. Idk what to do about that. It says in my task manager that is is 2933.

Your DDR4 memory speed is correct. It says 1500MHz for one stick, but in Dual Channel it's 2x1500MHz (you double the bandwidth by splitting the data across two sticks).

If you ran CPU-Z (from CPUID) it should show in the memory section that it is in "DUAL" channel and show almost 1500MHz for the frequency.

If it showed SINGLE channel that would be a problem.

 

[wiseoxejah]

You can also PM me if you want, however a response may not be timely.

No need to offer to pay. I'm not sure if that's even appropriate here.

So we need to know EXACTLY why you think your performance is lower in a repeatable situation. Preferably an offline game/benchmark to avoid networking as part of the equation.

I just ran userbenchmark and it appears to be my cpu. Results are here. I will run any game or benchmark you'd like. I'll do the Heaven one now. UPDATE: PerformanceTest says my ram is running at 1501.5mhz. Idk what to do about that. It says in my task manager that is is 2933.

Your DDR4 memory speed is correct. It says 1500MHz for one stick, but in Dual Channel it's 2x1500MHz (you double the bandwidth by splitting the data across two sticks).

If you ran CPU-Z (from CPUID) it should show in the memory section that it is in "DUAL" channel and show almost 1500MHz for the frequency.

If it showed SINGLE channel that would be a problem.

Ahh. Thanks

 

[wiseoxejah]

Something is really wrong with those results. The CPU should not be a bottleneck, so assuming you had the EXACT same settings your GPU must be getting stuck at a lower frequency.

Possibly at around 1200 to 1300MHz?

Try running the benchmark again whilst monitoring with MSI Afterburner, EVGA Precision OC or suitable tool that has an OSD to overlay BOTH of these:

1) GPU frequency, and
2) GPU usage

Normally in that benchmark you should see about 95% usage and 1900MHz+ frequency on the GPU (except between scene transitions).

You might have for example, 1250MHz frequency and high GPU usage. (You should get the same thing if the CPU was the bottleneck but it would have to be really bad to be a bottleneck for this benchmark.)

*If you manually overclocked, try doing this:
a) remove any software used to OC or monitor the graphics card
b) run DDU http://www.guru3d.com/files-details/display-driver-uninstaller-download.html
c) reinstall the video drivers
d) rerun the benchmark

if still wrong, then
e) verify CPU frequency under load with Task Manager ("CPU Speed" should say about 3.7GHz?) for say Prime95 (just enable half the threads)

(right-click START)-> TM-> Performance-> CPU-> "Speed" (it should vary a lot in idle usage, then be pretty stable under load)

f) if still wrong (I mean if CPU results are good), contact the graphics card manufacturer

I ran the benchmark again and the clock stayed at around 1924 but the usage was at 99%. If you are curious, it was only 67c so it wasn't thermal issues.

[strike]One thing I did find odd was that my score went up.
My results:
-75.1FPS avg
- 1893
- 21.2fps low
- 142.7fps high[/strike]

That was heaven. Oops.
-49.7FPS avg
- 2081
- 17.2fps low
- 96.9fps high

Also, I tested my cpu with prime95 and it maxed at 3.15ghz. I upped the clock to 3.7ghz before doing the above benchmark, but not much improvement.

 

[wiseoxejah]

I know this thread is old, but I have updates. I ended up buying a new mobo and installing a fresh OS. It worked temporarily. I got great performance on benchmarks and games. I eventually saw slightly lower performance and after installing a new cpu cooler, it's back to being horrible. My temps never go above 60 now with a 3.8ghz overclock, but my cpu-z score is now way below reference (it was 17% above reference previously). I'm also back to lower framerates, some games are nearly unplayable at ultra.