Difficulty to POST on B450 ITX/ac with 3900X

Has this been fixed by anyone yet? What seemed to work for me is cleaning my amd ryzen 2700x with alcohol for like a hour and so gar so good . It may just be a Coincidence but figured I'd mention it since I know how annoying this can be . How can a company put out a product that barely works and not get any backlash for it I've noticed this with all the motherboard manufacturers, they just keep getting away with sub par crap. You would think by now they would have a good understanding of how all this works but I guess I'm wrong. Maybe I will launch a motherboard that I'll advertise as the best thing ever then have it only work half the time.

PeterCxy wrote: OK, so after a day of testing, I can say that resetting my RAM to 2133MHz (by out-of-box default it is at 2400MHz) everything starts to work fine. There's no POST issues now. It seems that going anywhere beyond 2133MHz will cause breakage with the current revision of BIOS.

How does it feel to use your new Ryzen with default memory speed?

Have the same problem with my x370 Gaming itx board and 3700X :( After I go to BIOS, it gets into a boot loop without post. The way that work for me is to pool out the batery and boot with my 1700X, shut down and change to 3700X... :(

Rpeck wrote: Set your memory to bone stock, not XMP. I know it's not a permanent solution, but more like a test. A lot of people are having issues with stability, posting or booting the 3000 CPU on ASRock boards with memory at 3000+ speeds. Seems to be a BIOS or AGESA issue and so far these dipsh*ts haven't done anything to address it. I'm having the same problem on a Tachai X470. Even the firmware they just released 4 days ago was about 4 revisions old from AMDs recommend AGESA. Methinks these people may not be too bright.

My memory is already running at bone stock, DDR4-2400. Since it can't even POST consistently, I didn't even bother to load the XMP profile, instead I just confirmed that it's really running at bone stock. So the problem doesn't seem to be memory.

I have had a Fata1ty B450 Gaming ITX/ac board paired with a 2700X for several months, and now I decided to upgrade to 3900X. I upgraded the bios to P3.30 which is supposedly the one that introduced support for 3000 series, then ordered a 3900X and swapped it in, along with a CMOS clear.

However, this board seems to have difficulty POSTing with a 3900X (not sure if other 3000 series have the same problem). More specifically, the symptom is that sometimes the board gets stuck in a boot loop without POST. Upon pressing the power button, the board basically does nothing and the CPU fan does not spin at all. Then it just stops itself and keeps rebooting (I can tell it reboots because the noise the PSU makes when switching on / off).

I have tried waiting for nearly half an hour but it seems to be just stuck there without any progress. The only fix I now know is to completely unplug the whole thing, wait for a while, and maybe clear the CMOS, then re-plug and boot again. It will still boot loop once or twice after this, and then the CPU fan starts spinning but still without POST. At this point, I can just force-reboot it by long-pressing the power button and then it will start POSTing just fine. Problem fixed, and I can boot into OS, but then after some time if I try to boot the computer, this may happen again (extremely often), and I have to go through this entire process to get it to POST again.

I don't think this is a problem of my CPU, because once it boots, everything works just fine. Every connected device shows up, and torture tests can complete without errors. It's just that it cannot POST consistently, and it seems that it's not just me with this problem.

As a bonus, rebooting from OS doesn't seem to work either -- once OS shuts down, the computer just hangs with CPU fans at Max. A normal shutdown without reboot is completely fine. I don't think this is an OS problem, because the same thing happens with the "save and exit" option in BIOS. Normally this should reboot the system, but in this particular case it just exits and make the whole thing hang.

Also, the CPU temperature sensor is invisible in Linux. I don't know if this problem is Linux-specific, but I cannot see any CPU temperature sensors even if I inspect PCI/I2C buses directly, which doesn't sound like a missing driver. It seems to me that the sensor is simply not reported to the OS (BIOS reads the temps seemingly fine) This was definitely not the case when I had a 2700X. (Linux has already included support for Ryzen 3000 cputemp sensors since the end of 2018)