• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Post your Cinebench R23 Score

Not bad . Screenshot 2022-02-05 075950.png
Screenshot 2022-02-05 075950.png
 
Last edited:
A tiny score from a tiny CPU :D
 

Attachments

  • Capture.PNG
    Capture.PNG
    2.3 MB · Views: 175
I need help trouble shooting my 5600x single core score.
I should be over 1500? I am getting 1373 with PBO on.
My multicore is 11,200+ up from 10,200.
But the SC shows basically zero gains?
-
I am on the stock 80mm air cooler (~55c single core, ~86c multi core in CB 23 runs).
I see basically everyone getting better than 1500, even at stock clocks.
-
My curve optimizer is 5 on the two fastest and 25 on the rest.
+25 mhz boost override and 1x scalar.

PPT 110, TDC 65, EDC 110
These are the max values that CB 23 was demanding, so I set my limits there.
 

Attachments

  • wtf.jpg
    wtf.jpg
    180.5 KB · Views: 157
Last edited:
I give my 5600X almost the same power limits that I give my 5900X lol.. she loves it rough like that. So something like 200ppt, 140tdc, 180edc +200 -29 is my lowest. It’s probably not the proper way of doing things..
 
I just got my 5600x and running static 1.25v and 4.6ghz with no issues in p95, 4.8ghz gave a bsod and 4.7 failed workers on the second pass in p95 but 4.6ghz is working nicely.

cinebench1.jpg
 
I just got my 5600x and running static 1.25v and 4.6ghz with no issues in p95, 4.8ghz gave a bsod and 4.7 failed workers on the second pass in p95 but 4.6ghz is working nicely.

What cooler are you running?
 
I need help trouble shooting my 5600x single core score.
I should be over 1500? I am getting 1373 with PBO on.
My multicore is 11,200+ up from 10,200.
But the SC shows basically zero gains?
-
I am on the stock 80mm air cooler (~55c single core, ~86c multi core in CB 23 runs).
I see basically everyone getting better than 1500, even at stock clocks.
-
My curve optimizer is 5 on the two fastest and 25 on the rest.
+25 mhz boost override and 1x scalar.

PPT 110, TDC 65, EDC 110
These are the max values that CB 23 was demanding, so I set my limits there.
I've yet to figure out that single core problem myself...stuck at 1498.
 
I have to beat on mine to get it to do 4700

1 2.PNG
 
CoolingWraith Prism

That is the problem the wraith Prism while being a decent stock cooler is not enough

but this thread has some great tips on how to improve core clocks and perf while keeping your CPU cooler by not allowing to much Volts to be pumped into it

https://www.techpowerup.com/forums/...-3-chips-pbo-settings-temperature-fix.283776/
Had the best score with the Prism cooler...since have replaced it with a Noctua NH-D15 and it made no difference. There is something missing here besides temp and MHz that will end up being stupid simple. Freeagent has a SC score 1000 points higher with a lesser cpu...doesnt add up. Tabacosauz using the same 5700G scored a 1553 in SC, but did not come close to my MT score of 15500...just so confusing for me.

Finally...I'm done with this mess lol! :wtf:

Finally.jpg
 
Last edited by a moderator:
Had the best score with the Prism cooler...since have replaced it with a Noctua NH-D15 and it made no difference. There is something missing here besides temp and MHz that will end up being stupid simple. Freeagent has a SC score 1000 points higher with a lesser cpu...doesnt add up. Tabacosauz using the same 5700G scored a 1553 in SC, but did not come close to my MT score of 15500...just so confusing for me.

Finally...I'm done with this mess lol! :wtf:

I'm almost certain it's my motherboard not supplying consistent TDC/PPT
I crash at +50 mhz even with curve set to 0 or 1.
Never buy an MSI PRO series.
My 3600 overclocked like shit too.
 
I'm almost certain it's my motherboard not supplying consistent TDC/PPT
I crash at +50 mhz even with curve set to 0 or 1.
Never buy an MSI PRO series.
My 3600 overclocked like shit too.
I don't think anyone makes things like they once did Zogg. :shadedshu:
 
@rickss69 can't compare 5600X to our Cezanne APUs, the chiplet CPUs will win every time. Just how it is, I guess. 1500 is respectable. We have about the same effective clock in ST; run-to-run variance, Windows services, etc. takes care of the rest. You'll always beat me on MT, I use either a NH-L12 Ghost S1 or Big Shuriken 3 on my 5700G, way way way less capacity than your D15.

Here's a very lazy 4400 XMP run on my Impact under my C14S. Are you using Curve Optimizer?

5700g r23 mt st impact c14s.png
 
Last edited:
@rickss69 can't compare 5600X to our Cezanne APUs, the chiplet CPUs will win every time. Just how it is, I guess. 1500 is respectable. We have about the same effective clock in ST; run-to-run variance, Windows services, etc. takes care of the rest. You'll always beat me on MT, I use either a NH-L12 Ghost S1 or Big Shuriken 3 on my 5700G, way way way less capacity than your D15.

Here's a very lazy 4400 XMP run on my Impact under my C14S. Are you using Curve Optimizer?

View attachment 237091
My memory is not spectacular...unsure if that has any bearing. The 15500 MC run was with the Prism cooler, but yeah, the D-15 is certainly the way to go. Since I swapped out those ridiculous 140mm fans for 120's I don't mind using it at all. I did use the Curve Optimizer for the 1506 SC run though likely not as well as your settings.
 
My memory is not spectacular...unsure if that has any bearing. The 15500 MC run was with the Prism cooler, but yeah, the D-15 is certainly the way to go. Since I swapped out those ridiculous 140mm fans for 120's I don't mind using it at all. I did use the Curve Optimizer for the 1506 SC run though likely not as well as your settings.

Mine is about -10 for all the cores I think. Not really completely stable to the standard of my daily 5900X's curves, I know that for a fact, but good enough not to crash and I can't be bothered.

Memory usually doesn't matter at all for "Ryzen" but I'm not prepared to say that it doesn't matter for the APUs. I just haven't tested thoroughly enough to find out. We have a fair bit more in common with the Intel chips (no off-die Fabric links, small L3, etc.), and memory matters a little bit more for them on R23, so.

Might be able to eke out maybe 5 points tops ST if I set up my 4333CL16 profile, I don't expect any more.
 
Well, well, well gentlemen
Look who fixed his SC score!

@rickss69 - you were right it was something stupid simple.

I uninstalled windows 11, clean wiped the drive, went back to windows 10
Raised MC by 300 points and SC by 170

90 PPT, 65 TDC, 110 EDC
+25 mhz override, 1x scalar, stock cooler
4 - 4 - 24 curve

EDIT:
After installing AMD Chipset drivers?
I get 300 less MC and 200 less SC
When I uninstall them and restart? I get my full score.
-
The dilemma has been solved gentlemen!
The answer is to throw AMD builds in the trash.
 

Attachments

  • best.png
    best.png
    122.6 KB · Views: 131
Last edited:
Well, well, well gentlemen
Look who fixed his SC score!

@rickss69 - you were right it was something stupid simple.

I uninstalled windows 11, clean wiped the drive, went back to windows 10
Raised MC by 300 points and SC by 170

90 PPT, 65 TDC, 110 EDC
+25 mhz override, 1x scalar, stock cooler
4 - 4 - 24 curve

EDIT:
After installing AMD Chipset drivers?
I get 300 less MC and 200 less SC
When I uninstall them and restart? I get my full score.
-
The dilemma has been solved gentlemen!
The answer is to throw AMD builds in the trash.
Seems like AGESA version can affect performance as well. I also recommend using AGESA1.2.0.3c over 1.2.0.5

According to several feedbacks from the forum, the author's own tests with an AMD Ryzen 9 5950X (test) on an MSI MEG X570 Unify and several discussions on Reddit, installing AGESA ComboAM4v2 1.2.0.5 can lead to the following problems:

  • "Drops" of up to 100 MHz with the single core clock
  • "Drops" of up to 150 MHz with the multi-core clock
  • Missing overclocking functions eg with PBO
  • A significantly cropped CBS/PBO menu
  • Performance issues with the second CCX
  • Low fabric clock (FCLK) possible
  • More WHEA-19 errors after OC

 
I am on AGESA 1.2.0.5 -- it must conflict with the Chipset drivers, namely the powerplan.
-
I installed chipset drivers twice today, and saw degradation to my Cinebench score each time.
After uninstalling and restarting? My score went back to normal.
This isn't margin of error, its hundreds of points.
-
The third time I left the "AMD powerplan" out of the install.
I have maintained my scores all day since.
 
I don’t use their chipset drivers, just whatever windows has for updates. I can’t see it being a big deal.
 
@The King

Thank you for bringing the AGESA version to my attention.
Turns out MSI's beta bios (the only one available at the time that supported a 5600x)
.. was terribly buggy for this PRO-VDH MAX.
They've recently released a new official build that has solved - everything.
-
I can now hold 3600 mhz on my RAM
And get my boost clock to +100 mhz stable
.. had crashed on anything over +25 mhz
I'm even pulling a plus 20 FPS avg to my 1% lows in gaming somehow?
-
1561 single core up from 1373!
Kudos to this forum.
Thanks fellas.
 

Attachments

  • best.png
    best.png
    49.1 KB · Views: 99
Last edited:
Not my rig; I just built it for a friend and benched it:

1645566169953.png

Single core seems a little low but the multi seems exactly right.




These 12400F CPU's are pretty much the new value king in the UK as the equivalent AMD setup would be around £75($100) more.
 
Just a quick run with stock 1700X CPU Clocks and DDR-3800 CL16. :)
CINEBENCH_R23_CPU_Multi_Core_8918.jpg
 
Screenshot (42).png


Not my best, bet better than most for a 5600x
cooling: corsair h150i capellix
I'm still running memory slower than most people.
PBO right now 4,600mhz all core 4,850mhz single
 
Last edited:
The last 500 points are always the hardest :D
 
Back
Top