PC Processors

[Schweb]

Hey all.  New here.  Just built a new PC.  8600g I will include complete specs below.

Mobo: PRIME B650M-A AX6 II

CPU: Ryzen 5 8600g.

Ram: G.SKILL Flare X5 32GB (2 x 16GB) 288-Pin PC RAM DDR5 6000 (PC5 48000) Desktop Memory Model F5-6000J3038F16GX2-FX5

SSD: Western Digital 1TB WD Blue SN5000 NVMe SSD, PCIe Gen 4.0, up to 5,150 MB/s Read Speeds Internal Solid State Drive (SSD) 

Case: Fractal Design Pop Mini Silent Black mATX Sound Damped Solid Panel Tower Computer Case

COU Cooler: Thermal right Assassin King 120 SE CPU Air Cooler, 5 Heat pipes, TL-C12C PWM Fan

My Question is when I so 1 Cinebenchr23 run on Multicore I am getting a score of 1700.  All over the web I am seeing an average score of 14000.  On single core 1 pass I get 1700 as well which is the normal as everyone else.  Any idea why this would be happening?

[misterj]

Schweb, post a screenshot of Ryzen Master (RM) running Cinebench R24 multicore. There is an * next to your score that I am not familiar with and may mean something bad. Notice that your low score is the best in the list. Get rid of R23 and get a fresh copy of R24 here. You have changed something in Cinebench. Run it as is. John.

[Schweb]

Ok any screen on Ryzen master you want to see when it runs?  I can do it after work

[misterj]

Thanks, Schweb. RM Advanced view. Clear CMOS and no images of Cinebench. Make no changes to Cinebench. R24 scores will be much lower than R23. John.

[Schweb]

Ok clear cmos just pull battery off motherboard for 20 seconds? What do you mean by no images of cinebench?

[misterj]

Schweb, please do not post any screenshots of Cinebench, just report the score. Thanks, John.

[Schweb]

ok just screen shot of RM and clear CMOS. Reset EXPO on memory?

[misterj]

Schweb, do one of each, please. John.

[Schweb]

So I did what you said.  I ran r23 again as an experiment with Ryzen master.  Only one of the cores was working, Number 6 core, when doing the multicore test.  Once I put on on r24 everything changed.  All cores on Ryzen master woke up, the fan reved up and away she went.  Its still going through pass 2 now but so far is has a score of 776.  I feel there might be like abug or something with the amd drivers or something that is preventing all cores to work on r23

[Schweb]

Final score was multicore 788.  Single Core was 110.  According to cpu-monkey score they got 770 and 101 I believe.  This is a screen shot of RM doing single core run 

[misterj]

Schweb, looks like you are doing fine. John.

[Schweb]

Must be just a weird bug or something with that r23.  All my HWINFO and everything look good to you John?

[misterj]

Schweb, it is not an R23 bug. It is the fact you changed something in R23 or DLed a modified R23 (MP ratio is set to 1.00 and Minimum duration is set to Off). I am not aware of what these two settings do but the settings are the reason I insisted you go with R24 Directly DLed from Maxon. RM looks OK but I really need to see Multicore. I do not trust the others so spent no time looking at them. John.

[Schweb]

That was multicore...........

[misterj]

OK, Schweb. I thought the first SS was R23. You are throttling due to PPT which my processor does also and is probably why you are only running 4633 MHz instead of 5000 MHz. Ryzen slows the clock so the processor does not overheat. Still looks fine. John.

[Schweb]

Well John cause I love tech stuff and my mind wont rest till I fully understand it, also for S and Giggles, and cause I am bored, I re downloaded r23. You were right.  I must have touched a setting I just dont remember doing it and what I did.  I did 3 full runs on Multicore and got a score of 13939 which is on par with what I am seeing over the net.  Here are some screen shots of RM and HWINFO64 data WHILE running Multicore on R23.  Let me know what you think.RMTEMPS

[misterj · Accepted solution]

Schweb, looks about the same, all good. Maxon changed the program, so your processor throttled a little harder on PPT but all looks good. John.

[Schweb]

That was fun.  Lets build another and troubleshoot it. Thank you for the help.

[misterj]

Schweb, hope it runs out of the box but learn little that way. John.