AB350M Pro4 - Bios 2.50 - Issues with OC.

sobekite wrote: Hey all, I've ran into a bit of an issue while trying to overclock my Ryzen 1400 with the AB350M PRO4 (mATX, Bios 2.50) and was wondering if anyone had a solution or had a similar experience. Previously, I was able to overclock the 1400 to 3.7 at around 1.28v in the bios (non-updated bios, think p.1.0?). It only allowed me to set a desired frequency and didn't allow you to change the voltage, so it automatically set that voltage. This wasn't a very satisfactory experience, as I couldn't change anything else. (Note: It still worked!).  So I read up on the new bios update (2.50) and downloaded it. Once it updated, I was able to manually change voltage! Great! But the minute I try to change anything from auto in the voltages/frequency and then try to boot the screen stays black. I have to manually shut the system down and clear cmos from the jumper.  Oddly enough I was still able to overclock from Ryzen Master with (2.50), while I wasn't previously able to with the shipped default bios.  Anyhow, just wanted to see if anyone has experienced issues with overclocking from bios with 2.50. Am I missing something? I tried turning Cool' and Quiet off and then trying to set a normal and 'safe' OC, but it failed and just won't post.  Thanks! TLDR: With default bios (1.10) I was able to overclock from bios but not ryzen master. After updating to (2.50) motherboard won't accept any bios OC, no post black screen, but allows Ryzen Master OC'ing.

Odds are, by what you describe, it's memory related. Or, more likely BIOS related considering the platform is still maturing.

Please post you system specs with as much detail as you know.

sobekite wrote: I'm at work right now but let me see if I can't look up exact specs via model numbers. Also a small note: This motherboard has given a ton of temp/voltage monitoring errors, with CPU-Z showing voltages well over 2.4v on vcore .  HWmonitor will say one thing, CPU-Z another, AIDA64 entirely different.  .... Ram: Ballistix Sport LT Gray 8GB DDR4-2400 UDIMM x2 1.2v (16G ran in dual channel, confirmed in 2.50 bios) .... Is it possible to update back to the a previous bios? I'd give up voltage control just to not have to load up Ryzen Master everytime! Appreciate the help.

If you run more than one monitoring app at a time odd are this is what the errors are generated. The chip responsible does NOT like being accessed by more than one app. Been this way since... well, decades now. Os since I began my computing and building.

Please, when home post the Model of that memory please.

You can go back, but keep in mind you'll be forfeiting the newer AGESA releases rolled into the new BIOSes.

Edited by sobekite - 27 Jun 2017 at 3:46am

@wardog, thanks for the help. I will take your advice and update both CPU-Z and some of the other monitoring apps I've been using. Will get the memory model for you as well, although I am almost certain it's BLS8G4D240FSB. 

I realize that ram model isn't listed on the supported memory list on Asrocks support page, but it seemed to work without any issues. 

sobekite, with your memory speed at 2400 and the 2.50 UEFI version, if you do NOT OC the CPU, does the PC boot?

Did you try setting a CPU OC frequency, but leave the VCore on Auto, and does that boot?

Now that you can set the VCore, what voltage mode did you use? I should ask what voltage modes do you have, Offset, Fixed?

When you did the OC with Ryzen Master, what voltage did you set for the VCore? Or what VCore readin did you get in AIDA64?

In you screenshot picture, you can see CPU-Z 1.79 showing 2.56V, while AIDA64 shows 1.250V. That's an example of how that version of CPU-Z is wrong, since a VCore of 2.56V is crazy high. Note that the AIDA64 value of 1.25V, multiplied by two, is 2.5V. A VCore reading is data that must be translated into the correct value, and CPU-Z 1.79 was off by a factor of two for Ryzen processors.

It also looks like AIDA64 is only calculating VCore to two decimal places, if we trust CPU-Z 1.79 at all. My point is your VCore would be 1.258V.

Next, for an OC to 3.8GHz, a VCore of 1.253V would be on the low side, and if correct would indicate a very good Ryzen 5 1400 CPU.

So 1.253V at 3.8GHz was the Auto VCore value according to AIDA64?

Actually, what is the situation for your screenshot? Ryzen Master OC? The Auto VCore OC with the original UEFI version? Or something else?

You said you had a reading of 1.28V in the UEFI with the original UEFI version? I can tell you the VCore reading in the UEFI will be lower than will be needed in Windows. My Ryzen board's VCore in the UEFI is substantially lower than it is in Windows. But yours seemed to be higher in the UEFI, or almost the same as in Windows? Normally only one core of a CPU is active when in the UEFI/BIOS.