Question (Ryzen 3900x // Nvidia 2080 Super) New PC isn't performing as it should

Status
Not open for further replies.

tenorsax_

Commendable
Oct 25, 2017
16
1
1,515
Been struggling with this for a few weeks now so I figured I'd try to find help here.

My PC: https://pcpartpicker.com/list/BYWLyk

I built this PC about 2 weeks ago now. Every part is brand new except for the PSU, which I reused from my old PC. I mainly use it to play Destiny 2, what I'm told is a very CPU heavy game. With what I have, On ultra settings, I'm getting anywhere between 130-140 frames in low load areas down to 60-70 frames in high load areas. When I first installed it i was getting what I believe I should be getting: upwards of 200 frames at all times. Temps while playing are around 45-50 C and have never gone above that.
I've tried all the basic tricks and nothing has worked. I have all the latest drivers downloaded for all my slots, sockets, etc, and I'm not getting any closer to fixing it.
Any suggestions or help would be appreciated. Included are a few screenshots of my Ryzen Master and my task manager showing the utilization of my PC.

I am playing on 2 240 Hz 1080p monitors.

Thanks in Advance.

Task manager: View: https://imgur.com/a/rxrIdmC

Ryzen Master: View: https://imgur.com/a/L4JpsWC

Benchmark: https://www.userbenchmark.com/UserRun/23994421
 
Been struggling with this for a few weeks now so I figured I'd try to find help here.

My PC: https://pcpartpicker.com/list/BYWLyk

I built this PC about 2 weeks ago now. Every part is brand new except for the PSU, which I reused from my old PC. I mainly use it to play Destiny 2, what I'm told is a very CPU heavy game. With what I have, On ultra settings, I'm getting anywhere between 130-140 frames in low load areas down to 60-70 frames in high load areas. When I first installed it i was getting what I believe I should be getting: upwards of 200 frames at all times. Temps while playing are around 45-50 C and have never gone above that.
I've tried all the basic tricks and nothing has worked. I have all the latest drivers downloaded for all my slots, sockets, etc, and I'm not getting any closer to fixing it.
Any suggestions or help would be appreciated. Included are a few screenshots of my Ryzen Master and my task manager showing the utilization of my PC.

I am playing on 2 240 Hz 1080p monitors.

Thanks in Advance.

Task manager: View: https://imgur.com/a/rxrIdmC

Ryzen Master: View: https://imgur.com/a/L4JpsWC

Benchmark: https://www.userbenchmark.com/UserRun/23994421

How did you plug in your CPU cables? Does your motherboard have two 8 pins, or one 4 pin and one 2 pin?
 
possible thoughts (in no specific order):

  • bad GPU drivers
  • RAM is in wrong slots
  • windows power settings
  • your ingame settings are altered (see: resolution scaling), did you fiddle with nvidia optimization in the geforce panel?
 
Quick Update, bought and downloaded 3DMark since I'm told Real bench messes with AMD processors. This was my result. At this time, 3dMark's website appears to be down so I'm unable to compare this score.
 
possible thoughts (in no specific order):

  • bad GPU drivers
  • RAM is in wrong slots
  • windows power settings
  • your ingame settings are altered (see: resolution scaling), did you fiddle with nvidia optimization in the geforce panel?
  • My GPU drivers are the most recent NVIDIA drivers (Ver 441.87, released on 1/6/2020)
  • The RAM is in the correct slots
  • My power plan is currently on "High performance"
  • Ive tinkered a little bit with the settings in the Nvidia control panel to optimize the game. No resolution scaling.
 
  • Like
Reactions: throwawayaccnt
Open aida64 , inside it go to computer/sensors, scroll to the very bottom, find GPU 12V sensor (not just 12V, some GPU brands like gigabyte do not have that sensor though, not sure about MSI.) , and look at it on your second monitor while you game on your first one, if it's at 11.4~12.4 during gaming then it's not PSU most likely
 
While I don't know how well your CPU is supposed to score in 3DMark - Time Spy Benchmark, your GPU score seems pretty normal to me.

I just ran Time Spy test with stock settings both on CPU and GPU and my 2080 Super scores 73.15 FPS in Graphics Test 1, and 68.37 FPS in Graphics Test 2. My CPU score is slightly higher at 36,37 FPS. Total score is 11464

My GPU scores aren't much different than yours, and my card is factory clocked a bit higher than your MSI card which might explain the small difference.

But as low as 70 FPS at 1080p with your setup sounds very strange. I had to RMA my i9 9900K CPU, so I temporarily replaced it with an i3 9100F, and I had a 2070 Super at the time, and with AA Off I easily ran Destiny 2 at 100 - 110+ FPS @ 1440p (max settings, except shadows at medium and AA off). So something seems to be performing below expactations with your setup. Did you max out AA as well? That setting does have a big impact on performance.

Unfortunately I don't have any suggestions for a fix, but at least it seems your GPU scores are more or less "normal" (In 3DMark at least).
 
Last edited:
  • Like
Reactions: FullTank
Are you saying this because your CPU is failing the RealBench stress test? (downclocking or failing the hash)

No, I was just advised against using realbench by a friend of mine. According to him RealBench is known to mess with the results of AMD processors. Could be wrong but I'm trying everything
 
Ive been monitoring Ryzen Master while ive been playing D2, only thing that I notice is that the EDC is reaching into the mid 80 - low 90 percentages. Not sure if thats normal or not
 
No, I was just advised against using realbench by a friend of mine. According to him RealBench is known to mess with the results of AMD processors. Could be wrong but I'm trying everything
You were advised incorrectly.
The purpose of using RealBench stress test is to put a high stress load across your primary components. The only results you are looking for in RealBench is whether or not the stress testing completed successfully (no hash mismatch or failures). Otherwise you are looking at temps, frequencies, and voltages in other softwares like HWiNFO64, etc.
 
How does this translate to 'better'?

See Mark Kaine's reply

There are numerous discussions about it. Views on it vary, some say it makes a difference, some not. Thought id suggest it, can only try.
 
See Mark Kaine's reply

There are numerous discussions about it. Views on it vary, some say it makes a difference, some not. Thought id suggest it, can only try.

Interesting. I'm not sure what wizardry goes in to AMD's plan, but these test results look inconclusive. A couple percentage points is far from a slam dunk.
Logic tells me that it takes Windows/Intel some unit of time to ramp up the MHz speed in response to load. Yes, it's a minuscule amount, but it exists. I'll just keep my CPU locked at it's boost speed. 😉

Thanks for the link though. I really appreciate those that go out of their way to do all these tests - they are soooo time consuming.
 
What Ill try doing this weekend is formatting my drives and doing a fresh Windows install. Someone on reddit had a similar problem and a fresh Windows install worked for him

Otherwise I need to isolate the problem and send the part back
 
What Ill try doing this weekend is formatting my drives and doing a fresh Windows install. Someone on reddit had a similar problem and a fresh Windows install worked for him

Otherwise I need to isolate the problem and send the part back

Depends what the problem was for the person at Reddit. If Windows was carried over from another system then yeah a fresh install is best. Not reset, a full reinstall from usb. But you did this already yeah?
 
Status
Not open for further replies.