0

I upgraded my GTX 1050 ti to a GTX 1080 FTW in a prebuilt Dell XPS 8900. When upgrading the GPU, I replaced my PSU with a 750 watt PSU. I started getting intermittent stutters when gaming after these upgrades.

I looked at my temps and noticed my CPU temperatures were hitting 80-85 celcius under load. I thought this was my problem, so I swapped out my heatsink to a closed loop liquid cooler, swapped to a new case with more airflow and added an additional fan.

The intermittent lag is still happening after normal temps now, and I noticed my CPU is hitting 100% randomly with games, which wasn't happening prior to upgrading the GPU.

Other data points: I do stream on Twitch, but it was not causing issues prior to the upgrade. 16 GB DDR4 RAM.

Any help would be greatly appreciated. This one has me stumped.

Edit: As requested, here are some more details: PSU is a Corsair CX750m. The temperatures during 100% utilization are around 65° C after the addition of the liquid cooler. Graphics drivers are the newest GeForce Game Ready drivers. In-game settings are what GeForce Experience said was optimal. The stutter seems to happen when the spikes occur.

I thought I included it above, but my CPU is an i7 6700k.

Improve this question
7
  • 1
    Please post some detailed specs. What brand is the PSU ? What temperature is the CPU at when at 100% load ? What driver version do you use for the 1080 ?
    – Overmind
    Jul 26, 2018 at 6:03
  • 1
    Did you change the game settings? You might've been GPU limited before and now you're CPU limited. Do the 100% spikes accompany the stutter? Also check your streaming settings to use the proper encoder. What CPU do you have?
    – Seth
    Jul 26, 2018 at 6:40
  • I've updated the post with the requested information. Hope this helps. I wouldn't think I would he CPU limited since I'm running an i7 6700k.
    – Brock B.
    Jul 26, 2018 at 11:50
  • Now that you upgraded the graphics card, the CPU is probably the bottleneck for whatever you’re doing. Due to bugs or misconfiguration the CPU may not run at full frequency. Did you check that yet? HWiNFO64 can for example log both frequency, load percentage and all other relevant values.
    – Daniel B
    Jul 28, 2018 at 8:36
  • The latest bios update kills this machine, so dont update it dell.com/community/XPS-Desktops/XPS-8900-BIOS-2-2-1/td-p/…
    – JohnnyVegas
    Jul 28, 2018 at 9:00

4 Answers 4

Reset to default
1
+50

I recently installed a GTX 1050 ti in my pc, I noticed that I was getting some non network related stutters after the new install as well, and what helped me was double checking that my old graphics card driver is not conflicting with the Nvidia card driver. I'm not sure how you installed it but in the past I have had issues with this a few times, from A Radeon R7 install back in the day to this new Nvidia I recently put in..There are some tools out there that can help to see if there are any conflicting issues with drivers, however I would first check with device manager to make sure in there, and then you can look into additional fixes. Sometimes even if you have installed the newest driver, if the old one(meaning the old card is still registered in your device manager) is still showing up as enabled in your device manager try disabling it so that your newest card is the only GPU in operation. if that doesn't seem to be the case you could also try installing the latest drivers off of Nvidia website for your particular card and then doing a full restart to see if that helps. However I have a suspicion that the conflicting drivers could be the possible culprit here! In the rarest case you could try to use a tool like driver sweeper to completely remove the 1050 ti drivers, then do a full restart and see if that does the trick!

Improve this answer
2
  • I did run DDU after experiencing this problem to make sure I wiped all drivers and started fresh. I'm running the newest drivers provided by EVGA (which are also the newest from Nvidia). I will check device manager later, but I don't think the drivers are the culprit for me, sadly.
    – Brock B.
    Aug 2, 2018 at 18:00
  • I see, Well hopefully Device Manager is the culprit then, as that would be the "simple fix" Good luck figuring it all out though! Also just to double check make sure its not actually a network issue that your getting, sometimes you will get rubber-banding that isn't necessarily caused by your system but rather the servers on their end...id double check for ping spikes as well just to rule it out! In a rare case separate hardware drivers can sometimes conflict as well..I've seen a lot of weird scenarios working as a OS Engineer!
    – RickwhoPrograms
    Aug 2, 2018 at 19:19
1

I believe that this Nvidia tool will help you collect some more information on what is actually going on here:

http://www.nvidia.in/object/gpu-frame-capture-analysis-tool-in.html

From the Nvidia FCAT Description:

Measuring performance as it’s delivered on a monitor, FCAT identifies dropped frames, runt frames, micro-stuttering, and other problems that reduce the visible smoothness of the action on-screen, even when running at sixty frames per second and above as reported by FRAPS.

Unfortunately I'm unable to test this for you - I'm running an AMD card.

Any information you can glean from the tool would be much appreciated by myself and others trying to help you out.

Improve this answer
1
  • I'll give this a try. I haven't seen this tool before.
    – Brock B.
    Aug 1, 2018 at 13:54
1

Its possible that your monitor and your graphics aren't on the same refresh rate? (hz = fps ) which cause stuttering or screen tearing

your issue with the temperature might be because of your case is your gpu blocking airflow? is your cased designed to have a beast of a gpu inside it?

Improve this answer
1
  • While I agree that my monitor and GPU might not be on the same refresh rate and could get some tearing, what is happening here is not tearing. This is more of a freeze than a quick tear, and lasts around an entire second sometimes or more. As for temps, my temps are looking okay now. After swapping to a new case with better airflow and putting some better cooling in, it's been good.
    – Brock B.
    Aug 1, 2018 at 13:56
1

It may be that some codec(s) you have installed and is used by the games is not identifying the new GPU and so is using the CPU instead.

You could use Process Explorer, DLLs pane, to check all the DLLs opened by a game that is exhibiting this problem, and try to identify this codec.

You could also try to install a codec pack. My favorite is the K-Lite Codec Pack.

The codec might also be built-into some of the games, but in this case the number of solutions is limited to mitigating the heat problem :

  • Reducing maximum CPU usage in Power Options -> Change advanced power settings -> Processor power management -> Maximum processor state. Reducing the maximum by a few percents will be unnoticeable but will help in reducing the temperature of the CPU.

image

  • Improving the cooling
  • Avoiding these games.
Improve this answer
2
  • I think a portion of this might be right and has been what I've been suspecting all along. I feel like something is going on that is causing the game to run off of the CPU and not transfer enough of the load to the GPU.
    – Brock B.
    Aug 1, 2018 at 13:59
  • I have added some more advice.
    – harrymc
    Aug 4, 2018 at 9:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .