OC3D Cinebench R20 Benchmark Thread

WYP

News Guru
Cinebench R20 has been released, making now the perfect time for a good benchmarking session.

The rules are simple, run Cinebench R20 and report your scores with a screenshot and details about your system. Let's see how all our systems compare!

We will split things into two sections, Single-Threaded and Multi-threaded. This time there is no GPU section of the benchmark to worry about, it's all CPU with R20.

Report your Results like shown below

Score, CPU, CPU Clock Speed, Memory Speed (MHz), OS, Forum Name

Single-Threaded

316cb, Intel i7-5820K, Stock Clocks, 2133MHz DDR4, Windows 10, Wyp

Multi-Threaded

2176cb, Intel i7-5820K, Stock Clocks, 2133MHz DDR4, Windows 10, Wyp

Kb7E2eY.jpg


Single-Threaded Charts

1. 543cb, Intel i9-9900K, 5.2Ghz, 3400MHz DDR4, Windows 10, Ofan
2. 536cb, Intel i9-9900K, 5.1GHz, 3333MHz DDR4, Windows 10, Ofan
3. 508cb, Intel i9-9900K, 5Ghz, 2800MHz DDR4, Windows 10, FTLN
4. 507pts, Ryzen 9 3950X, Stock, 3600MHz DDR4, Windows 10, FTLN
5. 505cb, Ryzen Threadripper 3990X, DDR4, Windows 10, tinytomlogan
6. 487cb, Intel i7-8700K, 4800MHz, 3200MHz DDR4, Windows 10, looz
7. 485cb, Intel 8700k @ 4.7GHz, 3200MHz DDR4, Windows 10, Eddie long
8. 433cb, Threadripper 1920X, 4.2GHz, 3200MHz DDR4, Windows 10, Kei
9. 428cb, Threadripper 1920X, 4.075GHz, 3200MHz DDR4, Windows 10, Kei
10. 426cb, Intel i7-5820K, 4.49Ghz, 2666MHz DDR4, Windows 10, TheStoka
11. 422cb, Threadripper 1920X, Stock clocks ~ 4.15GHz PBR, 3200MHz DDR4, Windows 10, Arne Saknussemm
12. 400cb, Ryzen 1700, 4.0Ghz, 3200MHz 16gbDDR4, Windows 10, KingNosser
13. 397cb, Intel i7-6800K, 4GHz, 2400MHz DDR4, Windows 10, Wyp
14. 394cb, AMD Ryzen 5 1600, 3.95, 3000MHz DDR4, Windows 10, RobM
15. 384cb, i7 4790K, Stock Clocks, 1866MHz DDR3, Windows 10, Wraith
16. 375cb, Ryzen 1700, 3.8Ghz, 3200Mhz DDR4, Windows 10, Kingnosser
17. 359cb, AMD Ryzen 5 1600, Stock Clocks, 2133MHz DDR4, Windows 10, RobM
18. 351cb, AMD Ryzen 7 1700X, Stock, 3200MHz DDR4, Windows 10, Excalibur50
19. 316cb, Intel i7-5820K, Stock Clocks, 2133MHz DDR4, Windows 10, Wyp
20. 284cb, Xeon E5 2683 v3, 3.1GHz, 1100MHz DDR4, Windows 10, Thanos
21. 246, 10/20 Broadwell E ES Xeon, 2.3ghz, 2133mhz DDR4 (8gb), Windows 10, AlienALX
22. 239cb, i7 920, 3.8GHz, 1600MHz DDR3, Windows 7, notcool
23. 221cb, AMD FX-8320, 4375MHz, 1332MHz DDR3, Windows 10, tgrech
24. 201cb, Intel i5-2410M, Stock Clocks, 1333MHz DDR3, Windows 7, looz
25. 185cb, AMD FX-8320, Stock Clocks, 1333MHz DDR3, Windows 10, tgrech
26. 180cb, Xeon x5650, Stock, 1333MHz DDR3, Windows 7, Kei
27. 185cb, Intel Pentium G620, Stock, 1066MHz DDR2, Windows 10, ET3D
28. 116cb, Core 2 Duo E6420, Stock, 800MHz DDR2, Windows 10, ET3D
29. 82cb, Atom x7-Z8750, Stock, 2133MHz DDR4, Windows 10, ET3D
30. 49cb, AMD E-350, Stock, 1066MHz DDR3, Windows 10, ET3D

Multi-Threaded Charts

1. 30026cb, Ryzen Threadripper 3990X, DDR4, Windows 10, tinytomlogan
2. 19344cb, Threadripper 3970X, 4.425, 2133MHz DDR4, Windows 10, D.A.M IT!
3. 10270cb, 7980XE, 4.4ghz, 3733mhz DDR4, Windows 10, Kaapstad
4. 10057cb, AMD 3950X, 4.30GHz, 3200MHz DDR4, Windows 10 Pro, Dicehunter
5. 9889pts, Ryzen 9 3950X, 4.2, 3600MHz DDR4, Windows 10, FTLN
6. 6153cb, Threadripper 1920X, 4.1GHz, 3200MHz DDR4, Windows 10, Kei
7. 6099cb, Threadripper 1920X, 4.1GHz, 3200MHz DDR4, Windows 10, Arne Saknussemm
8. 6096cb, Threadripper 1920X, 4.075GHz, 3200MHz DDR4, Windows 10, Kei
9. 5572cb, Intel i9-9900K, 5.2Ghz, 3400MHz DDR4, Windows 10, Ofan
10. 5440cb, Intel i9-9900K, 5.1GHz, 3333MHz DDR4, Windows 10, Ofan
11. 5421cb, Intel i9-9900K, 5.1Ghz, 2666MHz DDR4, Windows 10, FTLN
12. 4697cb, Xeon E5 2683 v3, 3.0GHz, 1100MHz DDR4, Windows 10, Thanos
13. 3822cb, Intel i7-8700K, 4800MHz, 3200MHz DDR4, Windows 10, looz
14. 3776cb, Ryzen 1700, 4.0Ghz, 3200MHz 16gbDDR4, Windows 10, KingNosser
15. 3661cb, Intel 8700k @ 4.7GHz, 3200MHz DDR4, Windows 10, Eddie long
16. 3531cb, Ryzen 1700, 3.8Ghz, 3200Mhz DDR4, Windows 10, Kingnosser
17. 3326cb, AMD Ryzen 7 1700X, Stock, 3200MHz DDR4, Windows 10, Excalibur50
18. 3136cb, Intel i7-5820K, 4.49Ghz, 2666MHz DDR4, Windows 10, TheStoka
19. 3075, 10/20 Broadwell E ES Xeon, 2.3ghz, 2133mhz DDR4 (8gb), Windows 10, AlienALX
20. 2955cb, AMD Ryzen 5 1600, 3.95ghz, 3000MHz DDR4, Windows 10, RobM
21. 2837cb, Intel i7-6800K, 4GHz, 2400MHz DDR4, Windows 10, Wyp
22. 2813cb, Xeon x5650, Stock, 1333MHz DDR3, Windows 7, Kei
23. 2536cb, AMD Ryzen 5 1600, Stock Clocks, 3000MHz DDR4, Windows 10, RobM
24. 2176cb, Intel i7-5820K, Stock Clocks, 2133MHz DDR4, Windows 10, Wyp
25. 1861cb, i7 4790K, Stock Clocks, 1866MHz DDR3, Windows 10, Wraith
26. 1407cb, AMD FX-8320, 4375MHz, 1332MHz DDR3, Windows 10, tgrech
27. 1263cb, i7 920, 3.8GHz, 1600MHz DDR3, Windows 7, notcool
28. 1173cb, AMD FX-8320, Stock Clocks, 1333MHz DDR3, Windows 10, tgrech
29. 444cb, Intel i5-2410M, Stock Clocks, 1333MHz DDR3, Windows 7, looz
30. 356cb, Intel Pentium G620, Stock, 1066MHz DDR2, Windows 10, ET3D
31. 289cb, Atom x7-Z8750, Stock, 2133MHz DDR4, Windows 10, ET3D
32. 219cb, Core 2 Duo E6420, Stock, 800MHz DDR2, Windows 10, ET3D
33. 91cb, AMD E-350, Stock, 1066MHz DDR3, Windows 10, ET3D
 
Last edited:
Something wrong with this. Well, maybe there's something wrong with my PC.

Firstly my CPU is boosting to 2.3ghz, not 2.4ghz on all 14 cores (as it should) and secondly when running the single core benchmark my CPU is all over the place, ranging from 3.3ghz down to 1.1.

I'm also disappointed with the score, as it should be a bit better than it is but WCYD? At the end of the day I suspect that either the Spectre/Meltdown patches have nerfed my CPU (I would be more leaning toward that) or running different speed RAM is causing an issue, as I can't seem to baseclock my board at all now, nor enable OC Genie.

Either way it is what it is, no point crying over it.

suT2tOq.jpg


Single Threaded - 236
Multi Threaded - 3937

Edit. I should have the same/faster score than Mark. Doesn't make any sense at all. Single core boost is off.
 
Last edited:
Pretty lousy OC all things considered:

Single-Threaded

487cb, Intel i7-8700K, 4800MHz, 3200MHz DDR4, Windows 10, looz

Multi-Threaded

3822cb, Intel i7-8700K, 4800MHz, 3200MHz DDR4, Windows 10, looz


Screenshot
 
I will update the list in a bit, but I believe that the new CInebench uses AVX instructions, so boost clocks wouldn't be as high. AVX is a huge power hog, but it is one of the ways that the newer versions of Cinema 4D runs faster than its predecessor.
 
I will update the list in a bit, but I believe that the new CInebench uses AVX instructions, so boost clocks wouldn't be as high. AVX is a huge power hog, but it is one of the ways that the newer versions of Cinema 4D runs faster than its predecessor.

This is true but does not explain the chasm between a stock 5820k and my CPU in single thread score given my single core boost is supposed to be 3.1 ghz around dead even.

I'm also pretty certain it should not be throttling at 1.1ghz on single core. Heck, it runs them all at 2.3 under load ffs.

I'm wondering if it's the Windows Store somehow? Gawd knows. Way to alienate your users though.
 
You can download a portable version of it via TechPowerUp or Guru3D, if going through Windows Store bothers you.
 
Something wrong with this. Well, maybe there's something wrong with my PC.

Firstly my CPU is boosting to 2.3ghz, not 2.4ghz on all 14 cores (as it should) and secondly when running the single core benchmark my CPU is all over the place, ranging from 3.3ghz down to 1.1.

I'm also disappointed with the score, as it should be a bit better than it is but WCYD? At the end of the day I suspect that either the Spectre/Meltdown patches have nerfed my CPU (I would be more leaning toward that) or running different speed RAM is causing an issue, as I can't seem to baseclock my board at all now, nor enable OC Genie.

Either way it is what it is, no point crying over it.

suT2tOq.jpg


Single Threaded - 236
Multi Threaded - 3937

Edit. I should have the same/faster score than Mark. Doesn't make any sense at all. Single core boost is off.

Please use the same format as the other posters when listing your results. For example, I do not know your memory speed etc.

Pretty lousy OC all things considered:

Single-Threaded

487cb, Intel i7-8700K, 4800MHz, 3200MHz DDR4, Windows 10, looz

Multi-Threaded

3822cb, Intel i7-8700K, 4800MHz, 3200MHz DDR4, Windows 10, looz


Screenshot

Still miles ahead of my PC, though it is my office system. Will need to try the 1700X later to get some AMD in here.

This is true but does not explain the chasm between a stock 5820k and my CPU in single thread score given my single core boost is supposed to be 3.1 ghz around dead even.

I'm also pretty certain it should not be throttling at 1.1ghz on single core. Heck, it runs them all at 2.3 under load ffs.

I'm wondering if it's the Windows Store somehow? Gawd knows. Way to alienate your users though.

Shouldn't be. AVX could be a bigger power hog on older CPUs. Hard to know.

CPU scaling should be fine, as this bench is designed to work well with even AMD's newest 32 core CPUs. Could just be a quirk of older Xeons.

40x8zFr


Single-Threaded

535cb, Intel i9-9900K, 5.1Ghz, 3333MHz DDR4, Windows 10, Ofan

Multi-Threaded

5423cb, Intel i9-9900K, 5.1Ghz, 3333MHz DDR4, Windows 10, Ofan

https://imgur.com/40x8zFr

And we have our first chart-topper. Congrats. That will be a hard single-threaded score to beat.
 
What I have is a generation newer than the Haswell 5820k. It's Broadwell E, so it *should* have IPC around 5% higher than the 5820k, offsetting the 100mhz defacit I have. My RAM speed? in my sig fella, top link. 2133, which is why I didn't bother posting it.

My CPU won't overclock since adding four more DIMMs (2400mhz but running at 2133). Maybe that has something to do with it?

But yeah, when I got the CPU in June '17 I benched it and the results were the same on one core as my 5820k I replaced. So something is badly amiss when the CPU is clocking to 1.1ghz.
 
Yeah I can't install it on my laptop via the MS Store because the store says it is unable to read my RAM amount and so can't verify I meet the Minimum requirement of 4GB. Might have something to do with the fact it's still on the original October update that actually released in October as opposed to the 2019 "October" update which I can't actually update to via update. Might have to force the update with a pen drive and try again since I'm getting security warnings now. But two potato PC (FX8320 and i7 3612QM [35W Ivy Bridge Quad Core]) benchmarks on the way at some point (I'm interested to see how these two 2013 architectures stack up against modern parts, I might go back and try an OC on the FX8320 at some point, used to get an extra 1Ghz over stock with that part).
 
My RAM speed? in my sig fella, top link. 2133, which is why I didn't bother posting it.
I guess WYP's reason for the uniform format is that he can simply copy-paste the results to the leaderboard instead of having to do formatting for people as well.
 
You can download a portable version of it via TechPowerUp or Guru3D, if going through Windows Store bothers you.

Good to know. Hmm, Chrome marked the Guru3D one as a virus...

What I have is a generation newer than the Haswell 5820k. It's Broadwell E, so it *should* have IPC around 5% higher than the 5820k, offsetting the 100mhz defacit I have. My RAM speed? in my sig fella, top link. 2133, which is why I didn't bother posting it.

My CPU won't overclock since adding four more DIMMs (2400mhz but running at 2133). Maybe that has something to do with it?

But yeah, when I got the CPU in June '17 I benched it and the results were the same on one core as my 5820k I replaced. So something is badly amiss when the CPU is clocking to 1.1ghz.

As I said, use the correct format. See Looz and Ofan's posts.

Understand that this makes it a lot easier for me to update the charts, and it will waste a tonne of my time, in the long run, if I need to look through posts, sigs etc for specs.


Update -

I guess WYP's reason for the uniform format is that he can simply copy-paste the results to the leaderboard instead of having to do formatting for people as well.

Basically this. Huge time sink in the long run.
 
It's totally not working properly. I just opened up HWMON and it isn't using one core. It continually hops around, mostly at 1.1ghz. I went into my bios and I disabled the AVX offset yet still it won't load one core up at a time. I'm not doubting my main all core result, as I can see the CPU locked to 2.3ghz over all 14 cores and 28 threads, but in single core something is seriously wrong with this.

Maybe they will update and fix it? IDK. I have my machine up to date on the Windows side of things.
 
As I said, use the correct format. See Looz and Ofan's posts.

Understand that this makes it a lot easier for me to update the charts, and it will waste a tonne of my time, in the long run, if I need to look through posts, sigs etc for specs.

I don't want my score submitted so it doesn't matter. It's clearly totally wrong, and there is something wrong with the benchmark as my results defy logic. Like I said, my CPU should boost to 3.1ghz on one core meaning 100mhz slower than the 5820k meaning pretty much identical single threaded results. Like, exactly as I found them to be when I replaced my 5820k with this Xeon.

I *think* my multi core result is about right, as it's about 18% faster than a Ryzen 1800x @ 4.1ghz but the single core result is waffy AF.
 
Might be worth seeing how the CPU reacts to other AVX applications in single core mode, like the AVX enabled versions of Prime95.

On Broadwell-E there was also a Windows Update that messed up a few things. I had to delete a DLL file in windows to keep my overclock on X99.
 
Might be worth seeing how the CPU reacts to other AVX applications in single core mode, like the AVX enabled versions of Prime95.

On Broadwell-E there was also a Windows Update that messed up a few things. I had to delete a DLL file in windows to keep my overclock on X99.

How and what did you do dude?

Gaming seems fine. I am getting between 2.8 and 3.1 ghz meaning Metro Exodus at least uses more than 4 cores (that is when it drops to 2.8).
 
How and what did you do dude?

Gaming seems fine. I am getting between 2.8 and 3.1 ghz meaning Metro Exodus at least uses more than 4 cores (that is when it drops to 2.8).

I think using a new BIOS can fix it.

If that fails it requires changes to Win32. not 100% what the file is anymore, will need to check. Googling No Broadwell-E OC on Windows 10 should help.

Quick Update - The Standalone version of Cinebench R20 can be downloaded on Tech Power Up.

https://www.techpowerup.com/download/maxon-cinebench/
 
Well bugger me wid fish fork, there's only a bios about 5 months old for my board !

The funny thing is that would have been during the rebuild/recase/mod so I had absolutely no idea.

I shall have a poke shortly. It (MSI Live Update) has also found chipset drivers, usb drivers etc all newer than what I had. I would imagine all of it came out in the wake of Spectre/Meltdown whilst my PC and I were having a meltdown of our own :D
 
I've tested my "couch gaming PC" desktop here, all settings at BIOS default atm as it's just been rebuilt

Single-Threaded

185cb, AMD FX-8320, Stock Clocks, 1333MHz DDR3, Windows 10, tgrech

Multi-Threaded

1173cb, AMD FX-8320, Stock Clocks, 1333MHz DDR3, Windows 10, tgrech

(Sorry for image quality it's still mostly offline)
MwyNvcIzztUc_lONf2JLk56uh1eU8EXhbtHPJ8xbytNjNOnhp8uvT79V7KV5IFx5cy5yCifPzvDhtM1tJjCRsawIyAqwT8I_EU3h5Eo5GdeM6rcz4NqoCoXb2BAn5pZQ_7TBylfHimeq5prhccVNQoeQbrXfsfj_LrOAOrxMoEKk59CJr5jEfYN0jN3ZfbJ1mhdvoq9CPNI89kX5UUOMBW4TuAzSbCoQLyOhw--tKBSZOiU4eeiAFPsDn_fLop0AJ-mWnBm0SmJkfEsuSq1MMmuxrRqfsOx6dKcPW3avS5N56yu5aIdF6-SksSApBW2gPE10T1cKLVm1ZvkLKIjTmlIZAQN2VSMvYzxaaydzMbUMn-NMFktNFeJRt9mgpcSBhnfG0KC9t2Sqils26Okkrf1RGHm8qPcvtmbR6OhJAyBwlyu-RYBt_E4DyPKRRSaDkbMgx8bmZ-34JHn225Vg_b6LBdo1iohp_5wtpyAWqEVcsqvTYnjXarhdqpnatkt3KAR5ndndC9ryMMSLrUD03tLI6Hw0Ky0bdbD3mM3jsZtmJ1xNCe6bScUYAAZ9s7Yipb1MOsF5WofWGyTk0DQauwuWWP0Q9y1EnQcxqLsCgAmc8YtUiTavPapFC09ioQTr6F3n15rQ4KrRtqUBNsVRN7YUKZv3vpBS=w712-h949-no
 
Leaderboard updated. I wasn't expecting to see Piledriver on the charts! Nice one.

The single threaded test must have been a pain to sit through.
 
Back
Top