F5F Stay Refreshed Hardware Desktop Are there any issues with the iGPU in Ryzen?

Are there any issues with the iGPU in Ryzen?

Are there any issues with the iGPU in Ryzen?

T
TooMarlin
Member
60
01-07-2026, 09:52 AM
#1
Setup as described follows: ASRock B650M Pro RS, Ryzen 9 7900, RX 7600 XT, 64 GiB (2x32) Kingston FURY KF556C36BBE-32 (EXPO off, default timing profile), OS includes Linux, Windows 10, and Windows 11.
Issues appear only with the iGPU in GPU-accelerated applications, showing strange visual artifacts. The discrete GPU works fine, and memory tests with Memtest86+ show no errors.
I noticed that higher iGPU load reduces these artifacts, especially in games; clearing the dGPU, re-seating RAM, and testing different slots or connections didn’t resolve the problem.
It seems the issue might lie with the CPU, but I’m unsure before filing a warranty claim. I don’t have another AM5 test bench available.
Would you like suggestions for further troubleshooting, or should I consider sending the CPU for repair?
Thanks in advance.
T
TooMarlin
01-07-2026, 09:52 AM #1

Setup as described follows: ASRock B650M Pro RS, Ryzen 9 7900, RX 7600 XT, 64 GiB (2x32) Kingston FURY KF556C36BBE-32 (EXPO off, default timing profile), OS includes Linux, Windows 10, and Windows 11.
Issues appear only with the iGPU in GPU-accelerated applications, showing strange visual artifacts. The discrete GPU works fine, and memory tests with Memtest86+ show no errors.
I noticed that higher iGPU load reduces these artifacts, especially in games; clearing the dGPU, re-seating RAM, and testing different slots or connections didn’t resolve the problem.
It seems the issue might lie with the CPU, but I’m unsure before filing a warranty claim. I don’t have another AM5 test bench available.
Would you like suggestions for further troubleshooting, or should I consider sending the CPU for repair?
Thanks in advance.

I
Immotal_god
Member
122
01-07-2026, 09:52 AM
#2
I don’t see any other possible reason besides the processor. Consider verifying all processor voltages with hwinfo (vcore, vsoc, vigpu, etc.). It’s especially useful to test Vigpu under both load and idle conditions.
I
Immotal_god
01-07-2026, 09:52 AM #2

I don’t see any other possible reason besides the processor. Consider verifying all processor voltages with hwinfo (vcore, vsoc, vigpu, etc.). It’s especially useful to test Vigpu under both load and idle conditions.

N
Niroben
Junior Member
4
01-07-2026, 09:52 AM
#3
I would change those boards by updating their BIOS to the latest version. Older BIOS versions on ASRock boards were known to harm chips, so I don’t suggest them for AM5. Make sure your CPU has adequate cooling.
N
Niroben
01-07-2026, 09:52 AM #3

I would change those boards by updating their BIOS to the latest version. Older BIOS versions on ASRock boards were known to harm chips, so I don’t suggest them for AM5. Make sure your CPU has adequate cooling.

L
livvy66
Member
129
01-07-2026, 09:53 AM
#4
The images indicate the CPU might be heading toward failure. I’d think about initiating an RMA to check if AMD can replace it, and you should seriously look into switching to a different motherboard. If that’s not possible, ensure you’re running the most recent BIOS version. Then, with more conservative settings, adjust PBO power limits using a negative voltage curve as your next best approach to address the problem now or when you obtain a new CPU.

Only proceed with these steps if you feel comfortable modifying BIOS settings. Also, verify each change thoroughly.

Enable AMD overclocking first.
Given your DDR5 5600 RAM in a 2x32GB configuration, set the SoC voltage to 1.15v, and increase up to 1.18v if you plan to overclock to DDR5 6000.

In the PBO power limits menu, configure PPT to 142w/ TDC to 110A/ EDC to 150A. These adjustments will restrict the CPU’s power draw to approximately a 105W TDP from its 170W maximum. You may experience around a 5% performance drop or up to 10%, but stability should improve significantly.

In the Curve Optimizer menu, apply settings for all cores and set each core to a negative value of -15 (-15).

Alternatively, you might consider setting a thermal throttle temperature limit between 80–85°C, though this may not be necessary if your CPU cooler is adequate.
L
livvy66
01-07-2026, 09:53 AM #4

The images indicate the CPU might be heading toward failure. I’d think about initiating an RMA to check if AMD can replace it, and you should seriously look into switching to a different motherboard. If that’s not possible, ensure you’re running the most recent BIOS version. Then, with more conservative settings, adjust PBO power limits using a negative voltage curve as your next best approach to address the problem now or when you obtain a new CPU.

Only proceed with these steps if you feel comfortable modifying BIOS settings. Also, verify each change thoroughly.

Enable AMD overclocking first.
Given your DDR5 5600 RAM in a 2x32GB configuration, set the SoC voltage to 1.15v, and increase up to 1.18v if you plan to overclock to DDR5 6000.

In the PBO power limits menu, configure PPT to 142w/ TDC to 110A/ EDC to 150A. These adjustments will restrict the CPU’s power draw to approximately a 105W TDP from its 170W maximum. You may experience around a 5% performance drop or up to 10%, but stability should improve significantly.

In the Curve Optimizer menu, apply settings for all cores and set each core to a negative value of -15 (-15).

Alternatively, you might consider setting a thermal throttle temperature limit between 80–85°C, though this may not be necessary if your CPU cooler is adequate.

S
Sket
Junior Member
38
01-07-2026, 09:53 AM
#5
2 years using the same CPU and memory capacity showed no issues and did not harm the CPU. Problems only occurred with a limited number of MBs, not just Asrock chips, with 800 series chipsets and 9000 series Ryzen processors.
S
Sket
01-07-2026, 09:53 AM #5

2 years using the same CPU and memory capacity showed no issues and did not harm the CPU. Problems only occurred with a limited number of MBs, not just Asrock chips, with 800 series chipsets and 9000 series Ryzen processors.

N
NannaxI
Junior Member
20
01-07-2026, 09:53 AM
#6
It seems like something I should have addressed earlier might still be affecting me. I've been running this system for about six months now, and the iGPU problems started right away. I'm using the standard AMD cooler, but I'm concerned it might break the warranty.
N
NannaxI
01-07-2026, 09:53 AM #6

It seems like something I should have addressed earlier might still be affecting me. I've been running this system for about six months now, and the iGPU problems started right away. I'm using the standard AMD cooler, but I'm concerned it might break the warranty.

V
vdlogt254
Member
74
01-07-2026, 09:53 AM
#7
With PBO it shouldn't cancel the warranty. However, in my view, your issue is quite alike the VMIN shift problem, just for AMD iGPUs. I'm pretty sure raising the iGPU voltage or idle frequency could fix it, though that might not be the ideal fix.
V
vdlogt254
01-07-2026, 09:53 AM #7

With PBO it shouldn't cancel the warranty. However, in my view, your issue is quite alike the VMIN shift problem, just for AMD iGPUs. I'm pretty sure raising the iGPU voltage or idle frequency could fix it, though that might not be the ideal fix.

K
Keeszz
Junior Member
11
01-07-2026, 09:53 AM
#8
It might simply be a problem with a CPU pin not being in place, which could cause issues. Reattaching the CPU after confirming the socket might resolve the matter, but I recommend trying the PBO power limit first before disassembling.
K
Keeszz
01-07-2026, 09:53 AM #8

It might simply be a problem with a CPU pin not being in place, which could cause issues. Reattaching the CPU after confirming the socket might resolve the matter, but I recommend trying the PBO power limit first before disassembling.

I
iiSweeTzz
Posting Freak
862
01-07-2026, 09:53 AM
#9
The stock cooler looks pretty outdated if the issues started from the beginning, so I’d probably go for an aio with better performance instead. I’d also update the BIOS to the newest microcode.
I
iiSweeTzz
01-07-2026, 09:53 AM #9

The stock cooler looks pretty outdated if the issues started from the beginning, so I’d probably go for an aio with better performance instead. I’d also update the BIOS to the newest microcode.

L
LCAlex33c
Junior Member
35
01-07-2026, 09:53 AM
#10
I started by reseating everything after seeing the problem. Adjusting power limits and voltage curves improved stability a lot, so I’ll still forward the CPU for RMA after the holidays.
L
LCAlex33c
01-07-2026, 09:53 AM #10

I started by reseating everything after seeing the problem. Adjusting power limits and voltage curves improved stability a lot, so I’ll still forward the CPU for RMA after the holidays.