Is my 7900X faulty?

Nickhamm

I have built many Intel and AMD machine over the past 15 years. This week I built my first AM5 machine. Specs as follows:

ASUS B650E-F ROG Strix Gaming WiFi Motherboard

AMD Ryzen 9 7900X

G.Skill Flare X5 Series 32GB (2x16GB) DDR5-6000 RAM

Noctua NH-D15S Chromax Black CPU Cooler

WD Black SN770 1TB SSD

AMD Radeon RX 6950 XT GPU

Thermaltake Toughpower Grand RGB 850 Watt 80 Plus Gold PSU

Montech X3 Mesh Tempered Glass Case

I built the computer, putting the SSD in M.2_1 bay that's PCIx5.0 compatible (though the drive is a 4.0 drive, but I figured it's backwards compatible.) First boot posted, but upon running Windows 11 flash drive to install, it froze on the loading circle. Rebooted, BSOD "SYSTEM_SERVICE_EXCEPTION". Reboot, BSOD "IRQL_NOT_LESS_OR_EQUAL." Reseated everything. Turned on. Same BSODs. Took out front USB/Audio , SSD, 1xRAM. Got to Windows setup. Tried Linux, worked. Ran Memtest86, RAM passed. Got Windows installed on spare SATA SSD. Chipset drivers fine. GPU drivers BSOD while extracting "DPC_WATCHDOG_VIOLATION." In BIOS disabled Asus Performance Enhancements, CPU Core Boost, and PBO. GPU drivers install fine. PC benchmarks fine, though low scores. SSD back in bay M.2_1 and BSOD. Moved SSD to M.2_3 slot and tried again. Windows works and benchmarks fine. MB manual says M.2_1 is CPU controlled , M.2_3 chipset controlled. Enabling Core Boost, BSOD Windows. CPU temp in BIOS shows 36° when boost disabled, but if I enabled show 53°, both idle. Since the CPU can't seem to boost, the temps jump enabling core boost, and the M.2 slot it controls seems to not work, I think this is a faulty CPU. I am open to other thoughts and I only ask for additional advice or thoughts because I have never had a CPU go bad on me before.

TL;DR: New setup keeps BSOD or freezing, even loading Windows setup, unless I disable CPU Core Boost and don't use the CPU controlled M.2_1 slot. Is my CPU faulty?

PowerSpec_MikeW

@Nickhamm

I'd suggest trying the NVME in M.2_1 with the port set to Gen4 in the BIOS. Are you on the latest BIOS for the board with AGESA 1.0.0.5 Patch C?

Nickhamm

I am on BIOS 1222, released 2023/02/24. AGESA version ComboAM5PI 1.0.0.5 patch C. I do not see any options in the BIOS to change the M.2 mode. And that still doesn't solve the problem that I can't enable Core Boost or PBO.

PowerSpec_MikeW

@Nickhamm

Should be under the PCIe subsystem settings, unfortunately I don't have the board to check. I'd check ASUS enhancements disabled, PBO on. I wouldn't exclude the RAM at this point, try something that stresses the RAM. Memtest86 is an accuracy test rather than a stress test. I would repeat your testing under HCI Memtest, with the system boosting.

ricardo_ortega

Windows 11 requires TPM 2.0 enable in order to use windows 11. For some reason, AMD motherboards don't have it enable. Depending on the motherboard, you want to search how to enable TPM 2.0. That's the only thing I can see that it's not letting you boot windows. Since I had that problem too.

Nickhamm

@ricardo_ortega

TPM is enabled and working. My Windows install was straight from Microsoft's Media Creation tool with no modifications and will install, but only if I disable Core boost on the CPU.

@PowerSpec_MikeW

With Asus Enhancements off, PBO On, Core boost on, I cannot get Windows to stay booted long enough to even open HCI Memtest. If it doesn't freeze while loading, it loads to a desktop about 10 seconds before a BSOD. If I disable Core boost, back to stable again.

PowerSpec_MikeW

@Nickhamm

Test with a single stick of RAM installed on the JEDEC profile. If the results match that rules out RAM. Board or CPU, given the M.2_1 issue you discovered, CPU sounds like the better place to start.

Nickhamm

@PowerSpec_MikeW

I took out a stick of RAM, enabled Core Performance Boost, and got a Windows BSOD on load 3 reboots in a row. Went back into BIOS, turned it off, and it boots fine again and runs RAM test fine.

Also while I was in the BIOS, I checked into PCI Subsystem Settings. The only options I have are "Above 4G Decoding," "Resize Bar Support," and "SR-IOV Support." The first two are enabled and the last one is disabled. No M.2 options. I went through every menu I could and couldn't find any.

PowerSpec_MikeW

@Nickhamm

Don't have an ASUS AM5 board setup right now to check. It may not be a visible setting on the board/platform. You've done what you can, swap the CPU and if nothing else we'll rule out another component.

PowerSpec_MikeW

@Nickhamm

Please follow up with us and let us know if the CPU is an issue. I have just received a sample of this board and I'm testing with the 7900X and Flare-X 6000 C36 RAM. My experience so far has been that the RAM failed to POST on EXPO II. Did POST on EXPO I, however was unstable. Bricked the board on the first BIOS flash, had to recover via BIOS flashback. Testing on a Ripjaw 5600 C36 kit right now and beginning to put the board through our PowerSpec qualification process which likely will take 1-2 weeks. I will let you know if I'm able to replicate the M.2 issues you mentioned.

Nickhamm

I have a request in with AMD that is currently under review, so will do.

Nickhamm

I finally got my replacement CPU today and after installing it, Loading Optimized defaults, and reenabling EXPO, everything works great! I have had the machine up and running with CPU confirmed to be boosting above 5ghz and benchmarking in Geekbench above the average score for the 7900X. It's been running about 4 hours now with no crashes or issues.

I did goof and forgot to move the SSD before I installed the fan (that blocks access to the Gen5 SSD slot) but it isn't a Gen5 SSD anyway, so I'm still getting full speed on the Gen4 slot. I'll cross that bridge when I get a Gen5 SSD, but if it's like the rest of the issues, it won't be an issue now.

Again, thanks for your help. Hopefully your PowerSpec qualification process goes as well as my CPU swap did!

PowerSpec_MikeW

@Nickhamm

Glad you sorted it out. Bad IMC on the CPU. That's a rare failure.