Guys I have a problem with my new taichi x870e and my ryzen 9950x.
Similarly, I had a 9950X, Gen 5 SSD, GeForce 1030, and 192GB functioning in a ROG STRIX X670E-E. I transferred everything to my newly acquired X870E Taichi yesterday and was met with a series of unusual behaviors. My purchase better be defective, otherwise there's no justification for how poorly this turned out. Not even 24 hours after receiving it, the board is back in its container and in the hands of UPS. Instead of testing and tuning this weekend, I have to wait until the vendor receives my board and dispatches its replacement.
Right off the bat, the "MOS fan" was unnerving. Its motor made a noise that resembled what one might hear when an AIO pump is failing or straining. The noise continued when I removed the AIO wire from the header, so at least my AIO wasn't the problem. The low-pitched "buzzing" characteristic of the MOS fan sound just triggered my brain to think that something isn't right.
When I installed all of the aforementioned components, I did so with the intention of hearing a series of beeps (just to check how well it'd take 192GB up front) more than anything. Instead, while I watched the Dr. Debug go through all of its codes, the board simply shut off. Obviously, I had thermal paste and solid contact with my AIO, but that shutdown sure looked like a thermal runaway.
I decided to go down to 1 DIMM. Same problem. It wasn't until I removed the SSD and the dGPU that I finally got a POST and a solid beep. However, even then, I'd suddenly lose the USB port where my logitech keyboard dongle was inserted. The board detected the JEDEC settings from my DIMM, but was apparently so unstable that I could not even poke through the BIOS settings fast enough before the screen would black out or freeze. One on occasion, I managed to change the RAM speed to 3600 (as that's what the ASUS board liked) but the same weird shutdown behaviour appeared. The voltage and thermal readings all appeared to be nominal before saving and resetting. For this thing not to be able to POST properly with one stick at 3600?!? Wow.
The BIOS on my board was the very first release from July 2024. I had to use the flashback feature to get the latest version installed, because it was just not happening with "Instant Flash". This did not improve the issue with the SSD, however occasionally I could
almost POST when the dGPU was installed. With just the CPU and either 1, 2, or 4 DIMMs, and among different DIMMs (one of my 48GB Corsair sticks, and one 16GB 6000MT/s G.Skill) POSTing was maybe a 30% chance.
I plugged the power supply into a watt-meter to try to get an idea of what was going on. When the board would fail to POST, or get halfway and black out/freeze, I'd see the power level hang below 68W. Same thing would happen after attempting an Instant Flash (with the indicator being stuck at 0%). Meanwhile, when it came to those "shutdown" issues, I'd see the power consumption gradually rise from 70W, slowly through 80, 90, 100... and by 130-150W (and often near "00" on the Dr. Debug) that's the moment the board shut off. If I hit the power button again, the shutdown usually repeated. If I shut off the power supply and waited to power up, maybe 2 out of 5 times I could get it to POST or give me some other error response.
The board would either not POST or it would hang at the "Asrock" logo and go no further, if my Crucial T700 SSD was plugged in. Didn't matter what m.2 port.
Now I'm a little bit annoyed, but I'm gonna have to reassemble my ROG Strix X670E-E with the 9950X to make sure I didn't cause harm to the CPU and the other components that the Asrock did not seem to cooperate with.