I bought a Ryzen 3600x a couple months ago and I've seen this issue persist through all the bios agesa updates. At first I thought it was an underdeveloped boost algorithm or bios, you know, the early adoption 'fee'. No matter the work load, my CPU never reaches the advertised clock speed of 4.4 GHz, not even for a second.
I know I'm not limited by my cooling, and my motherboard should not limit it either as my temps stay under 65C and my motherboard is a MSI X470 "Gaming Plus".
The max clock speed I see on my best ccx is 3375 MHz and on my worst I see at max 3350MHz and these frequencies are not kept for long.
Is there anything I can do to get my cpu to perform as advertised or did I loss the silicon lottery and AMD should have made my CPU a non X skew.
Solved! Go to Solution.
Thanks for responding, my windows and bios are upto date and power plan is set correctly. My bios is correctly set with PBO enabled and the voltage, multiplier, and base speed set to auto. As far as I could tell I could not see a forum on MSI's website about similar issues.
That's odd then since it's not even running at the BASE frequency of the 3600X, which is 3800 mhz.
Do this. Open Ryzen Master, and while it's open run Cinebench R23 both single core and multi core, and monitor the frequencies. There's no need to let it run through its entire thing, a few seconds is all you need. What you should see as normal behavior is, under single core, one core running at its maximum speed while the rest clock down (top), and under multicore all cores will run as fast as possible as heat and power restrictions allow (bottom).
Post screenshots like mine below under those situations.
Thanks for responding and sorry for taking so long to reply.
Here are the screenshots of Ryzen master while running Cinabench r23 multi and single core respectively. For multi-core test, I observed my CPU frequency basically lock at 4.150 GHz without going over the allocated power or temperature limit. For the single core tests, I observed my CPU's active core bouncing around between core 1 and 2 (if cores index start at 0) of ccx0. The frequency of the active core was also jumping around between ≈4.000-4.325 GHz and ONCE it went up to 4400 GHz for very little time.