Hello Community,
I completed a 1080p build and I'm getting artifacts on second monitor while gaming. I noticed that frequency of artifacts is increasing when GPU is over 80C temp.
List of parts:
- In Win Chopin MAX with 200W 80+ Gold PSU
- Ryzen 7 8700G
- Kingston Fury Renegade Silver 6400 2x16 (running in XMP 6400MT/s)
- Gigabyte A620I AX, bios F22
- Noctua NH-L9a/AM5
- TG Kryonaut thermal compound
- Kingston NV2 1TB
I'm running latest drivers, including iGPU (installed new version with DDU), Windows is freshly installed and I use this machine only for MMO Gaming.
Artifacts are colored dots all over monitor #2 (DP 1.2, from mobo port). On first monitor, while gaming in full screen windowed there aren't, this is connected on HDMI port of mobo (#1 monitor)
Game played: FFXIV Endwalker 1080p with super resolution to 1440p. Temps are ranging on both CPU and iGPU between 79C and 85C (doesn't go over 85C even after many hours).
I ordered a new Cable Creation DP cable, as the one I'm using is a cheap one (HAMA), will update the post if the artifacts still occur after change.
What are the chances that my APU is defective?
Thank you.
Solved! Go to Solution.
Overheating GPU can cause artifacts to occur so can a Graphics driver or PSU power issues.
But your APU Maximum Operating Temperature is 95c so overheating is not the issue.
You have a 200 Watt PSU which is a very low wattage PSU. It may not be able to properly support dual Displays.
If the artifacts doesn't occur with just one Monitor but only when you have both monitors connected does seem to suggest a PSU issue.
With one monitor connected, download OCCT and run it three tests, especially the PSU Test and check for Temperatures, Fan speeds, and PSU Outputs (3.3/5.0/12.0 Vdc)
Now connected the second monitor and rerun the OCCT Test and check the PSU Outputs. They all should be within 5% tolerances. If the 12Vdc is below 11.4 that is indicating your PSU is not strong enough to run Multiple displays a the same time.
Overheating GPU can cause artifacts to occur so can a Graphics driver or PSU power issues.
But your APU Maximum Operating Temperature is 95c so overheating is not the issue.
You have a 200 Watt PSU which is a very low wattage PSU. It may not be able to properly support dual Displays.
If the artifacts doesn't occur with just one Monitor but only when you have both monitors connected does seem to suggest a PSU issue.
With one monitor connected, download OCCT and run it three tests, especially the PSU Test and check for Temperatures, Fan speeds, and PSU Outputs (3.3/5.0/12.0 Vdc)
Now connected the second monitor and rerun the OCCT Test and check the PSU Outputs. They all should be within 5% tolerances. If the 12Vdc is below 11.4 that is indicating your PSU is not strong enough to run Multiple displays a the same time.
The system is consuming only around 134W-135W with max stress on components, I used a smart plug to check and PSU is 80+ gold. There are no artifacts in synthetic benchmarks, only in gaming. Is either cable or heat, I will check after Cable Creation DP cable arrives.
@fxnoob wrote:I completed a 1080p build and I'm getting artifacts on second monitor while gaming. I noticed that frequency of artifacts is increasing when GPU is over 80C temp.
Artifacts are colored dots all over monitor #2 (DP 1.2, from mobo port). On first monitor, while gaming in full screen windowed there aren't, this is connected on HDMI port of mobo (#1 monitor)
Game played: FFXIV Endwalker 1080p with super resolution to 1440p. Temps are ranging on both CPU and iGPU between 79C and 85C (doesn't go over 85C even after many hours).
Artifacting can be caused by temperature but also too high memory clock, and 6400 is close to the limit. Have you tried disabling XMP?
Otherwise there are too many variables mentioned to isolate the issue. Temperature over 80°C. Full screen vs Windowed. FSR enabled and one particular game title.
Use something that will stress the GPU and is consistent/repeatable like Furmark. Start with one monitor to get a baseline, then add the second monitor and run again. Then switch Furmark between full screen and windowed mode. Change one thing at a time and you'll find the one thing that causes the issue. But I would start with disabling XMP first.
Disabling XMP is out of the question. All components are compatible one to another, I don't get any errors memory related, except artifacts. In synthetic tests I don't get any artifacts even APU is running at 95C, only during gaming. I will use a quality DP cable see if the problem is still there, the cheap ones may cause problems, including artifacts. And like elstaci said, temp is under TJMax.
It takes a few seconds to try just to see if it makes a difference, not saying to leave it disabled. The 8700G memory specification is 5200, anything over that is no guarantee. Most AMD IMC can run 6000 no problem, 6400 may be a stretch.
I disabled XMP and artifacts are still present during gaming. Two causes remain now, cable or heat. If is heat there is a bigger problem, because this CPU has TJMax of 95C, should work fine even at 95C, it protects itself anyway. If is cable, this I can only check after the high quality cable arrives, 2-3 days. I will post further after I troubleshoot everything.
Good to rule that out. Furmark stresses a GPU beyond any heat or load that a game would see, so if it doesn't artifact in Furmark then the problem lies elsewhere.
I went to Kingston website and inputted your Gigabyte Motherboard for compatible RAM Memory.
The highest for the Kingston Fury series was 6000 MTS: https://www.kingston.com/en/memory/search/model/108081/gigabyte-a620i-ax-motherboard?speed=6000mt%2F...
So @FunkZ might have a good point about running your RAM at 6400 Mts. Also check the QVL List and see if your RAM memory is listed the list for RAM. It does show 4 Ram Part Numbers running at 6400 Mts compatible with XMP but not EXPO.
I already stated that artifacts are still present even with EXPO disabled, running native 4800MT/s. They are less frequent, as heat is building up slower. From my experience using Gigabyte and MSI mobos, memory problems tend to be 0, Ryzen platform is already mature. So is either plain heat, or DP cable is causing artifacts as is heating up (more frequent after 80C).
My reply came at the same time your were writing your reply to the other User so I wasn't aware of that fact when I replied.
I would still run OCCT PSU Test just to see what your PSU outputs are like and to check your APU temperatures and fan speeds.
Also run OCCT CPU & GPU tests to see if any errors shows up with two displays connected.
OCCT is free and it doesn't require you to install in your PC.
Went to your PC Website and under specs it shows one USB-C output.
So you much have a splitter installed to connect two displays to one USB-C output?
I used GPU-Z and CPU-Z stress test integrated in them. I noticed that in fact there are artifacts, I wasn't paying attention before. With both running, only the PC without monitors is consuming 134-136W is way under the PSU's wattage, which is 200W and In Win is using certified Gold PSU in their micro cases. I'm using native ports that are in the A620I AX mobo, one HDMI and one DP. I have bad experiences with Type-C video signal adapters, even if they are top quality like CableCreation.
The reason why I mentioned OCCT because in its PSU test it runs both the CPU and GPU Test at the same time putting maximum stress on the PSU plus it shows all the data like Temps, fan speed, wattage, PSU Outputs, etc.
I realize that a low wattage PSU is sufficient for running a PC with just Integrated Graphics. But since you are using two monitors it might be putting stress on the 12 volt rail of your PSU.
I was trying to find the specs for a 200 watt PSU to find out what is the average wattage for its 12 volt rail but was unsuccessful.
If all troubleshooting goes to hell, I will use a Fractal Design case, like ERA ITX or Torrent Nano and I will be able to use higher wattage PSU and cooling. The problem here is my wife, the way she will react at using desktops again, after I used mini PC's a long time :). Will keep you updated how it goes.
forgot to post AMD SUPPORT link.
Open a AMD SUPPORT - WARRANTY ticket and see if they believe your APU needs to be RMAed to be checked or replaced from here: https://www.amd.com/en/forms/contact-us/support.html
To rule out a defective processor.
If you personally purchased from a Online/Brick Store an official AMD RETAIL BOX processor and the Retailer's Warranty or Return date has expired then your 3 year AMD Warranty goes into effect from date of Purchase.
Hello,
Update on situation.
Used CableCreation cables on both HDMI and DP and artifacts are still here.
Opened ticket to both AMD and GigaByte, gb's reply was a non professional one and with AMD I provided dxdiag and sysinfo files.
Gigabyte reply was something "on this low end mobo, you cannot game", weird that is almost 200 euro mobo, but I was expecting this kind of reply from them.
Could be what elstaci said, vrm problem and 12V rail overload as system is consuming more power and cannot cope up with dual monitors, if the CPU is under 60C artifacts are not there.
I'm waiting for AMD's feedback.
I found the cause of artifacts, a part of solution is from GigaByte and elstaci.
VRM's are too low spec for this load, even this mobo has 2 video out ports doesn't support 2 monitors + gaming, you have to use a 8+2+1 motherboard. With only one monitor active, with same setup on game (1080p upscale to 1440p), there are 0 (zero artifacts) even with 85C temp, so heat is not the problem here, is 10C below TJMax anyway and APU protects itself, plus a part of heat is going to case, being full metal.
Conclusion: for this type of build, if you use 2 monitors, gaming and a lot of load, don't go cheap on motherboard, get a high quality one even is 100+ euros more expensive.
I will change the mobo to a B650I Aorus Ultra in the near future.
Thank you for your advices and support.
The cause of artifacts was the overload of 12V rail from PSU (it was too weak). Elstaci figured it out :).
After rebuild in FD Torrent Nano, Seasonic 80+ Gold 550W + Noctua NH-U12S, artifacts are gone.
My APU is fine and working as is meant to be.
Gaming is 58-60C CPU, 55-57C iGPU. Synthetic stress CPU-Z and GPU-Z max 66C cpu, 54C iGPU. Idle in desktop CPU 38-40C, iGPU 31-33C. Max temp during Cinebench multicore - 80C.