PC Processors

[mmf]

I cannot get my new Ryzen 3600x to run stable on the hardware I used with the 2600 without problems, ASRock X470 Taichi and 3200C14 Trident-z memory. Are others having the same problem?

A memory clock far below XMP is needed to prevent WHEA_UNCORRECTABLE_ERROR or hard resets. The system even freezes occasionally in the BIOS and during a windows reinstall.

Maybe the memory controller has a defect but I can complete Cinebench runs with a good boost of up to 4.2-4.4 Ghz.
Things I have tried: Increased soc and dram voltage, manual and auto clocks, increased primary and secondary timings, different VDDP voltage, disconnecting non-essential hardware, windows re-installs, latest AMD X470 chipset drivers from 7/7, ASRock bios 3.40 and 3.43, re-seat CPU, re-seat GPU.
The cpu is cooled by water with a 420mm 6-fan radiator.

System: ASRock X470 Taichi, G.Skill Trident-z 3200c14 2x8GB, Geforce 1070 GTX, ASRock bios 3.40 and 3.43, Windows 10 pro x64 clean install, Samsung 970 Evo plus nvme.

[benman2785]

what is your soc and dram voltage?

i run 3400CL14 with 1.08V soc + 1.46V Ramvoltage

[mmf]

Also on Ryzen 3000 and an ASRock x470 mainboard?
I have tried dram voltage of 1.35V up to 1.55V. Soc is on auto around 1.0-1.1, but i have tried up to 1.25.
Even on default 2133Mhz I get hard resets while doing stress tests in windows. Also incompatible nvidia HD audio, graphics or PCI-e drivers could be another cause for this, but even navigating the BIOS has frozen several times.

I am returning the cpu as it's incompatible with my ASRock Taichi and Trident-z memory so far, perhaps with more driver updates it could work but this is unusable.

The same memory runs on 3466Mhz c14 on 1.43V stable on the same mainboard with Ryzen 2600.

[benman2785]

Ah - on my GA-AX370-Gaming 5 i needed to run a firmware flash tool before i should flash to the latest Ryzen 3000 bios - maybe AsRock has something similar to be done?

because your voltages should be high enough to ensure 3200CL14

[mmf]

Thanks for the tip. I am working with ASRock support now to see if a reversion to older BIOS and re-flash to Ryzen 3000 BIOS can fix this.
I'm now back on the Ryzen 2600 and all is stable again on the same BIOS at higher than XMP settings so it could be a bad cpu.
I'll post an update if this gets fixed.

[mmf]

A new ASRock bios with AGESA 1.0.0.3 has fixed this for me.