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

Post your Cinebench R23 Score

Devon68 said:
Here are my Cinebench R23 scores
This is my desktop running an FX 6100 at stock 3.3 GHz with 3.6 GHz Turbo (Mostly stays at 3.3 GHz while rendering). Air Cooled.
View attachment 175800
and this is my HP Laptop running an i7 7700HQ at stock (Seems to be running from 3 up to 3.6 Ghz, but mostly stays around 3.4 Ghz). Also air cooled.
View attachment 175801
Click to expand...
The max all core speed on 7700hq will be 3.4ghz.
 
@Nuckles56

well doesn't that just suck the lama's ball sack
 
Ryzen 9 3900 @ 44.25
32GB DDR4 3733
Air cooling, nH-D15
Screenshot 2020-11-15 154402.png
Screenshot 2020-11-15 155829.png
Screenshot 2020-11-15 160449.png
 

Attachments

  • Screenshot 2020-11-15 154938.png
    Screenshot 2020-11-15 154938.png
    102.6 KB · Views: 294
Last edited:
ASUS B450-E Gaming
AMD Ryzen 5 3600X, Stock
Air Deep Cool Neptwin V2, 3-FAN (CPU + CPU OPT + CHASSIS)
32G 4x8G DDR4 3600Mhz (OC) 16-19-21-36 (Hynix CJR)
55-56c - single thread, 70-71c - multithread.

cinebenchR23.JPG
 
arabus said:
This one better multiscore but why?
Click to expand...

You running Steam by chance? I had about a 15% drop awhile back that I discovered was being caused by Steam updating some apps in the background.. I'm sure it's not just steam that could be causing it, but likely some background process if you didn't change any settings...
 
Copied from post #1794 form @birdie
Cinebench R23
This bench takes forever to complete by default. To fix it: File -> Advanced Benchmark -> Minimum Test Duration: Off
Running it first on my laptop I got 3883 points for multicore, but after turning on the option above I got 4172 points. Might have been a fluke but some people should try it and se if their score improves as well.
 
Devon68 said:
Copied from post #1794 form @birdie
Cinebench R23
This bench takes forever to complete by default. To fix it: File -> Advanced Benchmark -> Minimum Test Duration: Off
Running it first on my laptop I got 3883 points for multicore, but after turning on the option above I got 4172 points. Might have been a fluke but some people should try it and se if their score improves as well.
Click to expand...
the score is greatly improved if the operating system is optimized properly. But I'm afraid a lot of people don't even know what it is.
Basic windows ran the test R20 4690 multi.
After optimization, it is 5000.
I guess that's why they didn't believe the results of the r23 base which was NOT OC with 12800 points.
Basically, a lot of 10700k doesn't even come close to this.
and that's it stock for me.
They later transcribed the base 3800mhz stock value to 5000 OC away from the board on the first page.
true later the result was also improved from 12,800 when i actually posted the result on 5ghz. :(
 
Last edited:
Semrpon 140 on AM2+
217sempronle140.png


I had horrible time getting this start. :-/ I had to use the Tab button and enter to start.
clicking on it wouldn't start it. Mouse is working fine though.¯\_(ツ)_/¯
Anyways 2.7ghz Sempron 140
I scored an an amazing 217 This test didn't take as long as R20 did on the same CPU.
It took maybe just over an hour.
R20 on the same cpu took over 2 /12 hours.
 
5600X Quick and dirty OC on 120 mm AIO.

12k should be possible on 5600X then (without extreme cooling).

cinebench_116.png
 
@T4C Fantasy The submission rules do not specify whether the test duration shall be enabled or disabled. I think it should.
 
FinneousPJ said:
@T4C Fantasy The submission rules do not specify whether the test duration shall be enabled or disabled. I think it should.
Click to expand...
I agree. In this day and age where power management on CPUs is rather complicated, running tests for a reasonable duration makes sure you're getting steady state performance numbers and not just numbers that reflect short duration boost. For the first 10 seconds or so, my CPU will do 4.5-4.6GHz on all cores, then it's down to about 3.2GHz if temps stay in check. Also it makes sure that your cooling solution can actually drive the clocks you're pushing as well. These are all good things IMHO.
 
Aquinus said:
I agree. In this day and age where power management on CPUs is rather complicated, running tests for a reasonable duration makes sure you're getting steady state performance numbers and not just numbers that reflect short duration boost. For the first 10 seconds or so, my CPU will do 4.5-4.6GHz on all cores, then it's down to about 3.2GHz if temps stay in check. Also it makes sure that your cooling solution can actually drive the clocks you're pushing as well. These are all good things IMHO.
Click to expand...
depends on the point of this thread...

To me, benchmarks are drag races...fastest from a to b...doesnt matter if it will make it across the state (is stable for longer).

This benchmark, the MT part, lives in a processors turbo boost duration anyway...at least for Intel. Amd is quirky...

But yeah, benchmark threads like this are for racing IMO, not to iron out steady state results and stability. Stsble is passing the test and getting a screenshot. That said, mine is steady state since its overclocked...as are many results here so. Im assuming they are steady clocks and not using turbo...though again amd and pbo, etc
 
Is that what the asterisk means; Minimum Test Duration <off>?

I tried both:

Screenshot (5).png
Screenshot (6).png


137 points difference. ~1%.
 
i will allow both to get the most scores as possible, not everyone is going to wait 10 minutes and some will, besides you can see who did what in their screenshots
 
I need to do more testing, need outside air temperature @20°F.
 
So, just to rule out the dGPU constraining thermal and power limits, I ran it one more time without the 5ks plugged in, and again for just a single pass. Single pass is an easy 500 points higher for me compared to my original multi-threaded run, probably because I'm thermally constrained on the laptop and because the VRMs run toasty and I haven't done anything to handle it yet.

Anyways this is the >10m run:
Screen Shot 2020-11-16 at 6.15.20 PM.png

This is the single pass run:
Screen Shot 2020-11-16 at 6.18.12 PM.png
 
Last edited:
Arctucas said:
I need to do more testing, need outside air temperature @20°F.
Click to expand...
with my 9900K doing the same test over and over i went from 12.8k to 13.2k bouncing back and forth, so its not like this test is strict
 
Last edited:
the 10 min multicore test is the average of all the times your CPU completes the test run which can depend on alot of things not just outright core speed it's also cache speed and instruction/data retention aswell

where as the zero time limit is a one off fast as you can do it run
 
biffzinker said:
It's 0°F where I'm at.
Click to expand...
Wednesday night has a low of 11°F in my neck of the woods. It's about 40°F right now. Since I stopped using the tower and switched to the laptop, I'm noticing myself getting colder more often in my office. :laugh:

Edit: Wednesday morning, I should go into the attic with the laptop and re-run the benchmark. See what really low ambient temps do for my overpowered laptop. FOR SCIENCE!
 
Personally regarding long term endurance (stability) vs. drag racing as applied to benchmarks, I do a lot of crypto mining with my rigs, so when I OC I go for the highest clocks I can get on the lowest voltage I can feed the CPU. That way it'll generate plenty of coin and run cool for days. Takes a good bit of trial and error any time I upgrade but my systems will go for as long as I can test them without a hiccup or coming remotely close to overheating.. With the R9 I've been able to run an all-core OC of around 4.1GHz at 1.15v under full load which is acceptable for my needs. I guess you could say it's drag racing vs formula 1.
 
I broke 12k on a single run. 5600X still on 120 mm AIO.

cinebench_120.png
 
You must log in or register to reply here.
Back
Top