Post your Cinebench R23 Score

[frostybe3r]

Intel Xeon W-3175X @4.5GHz AVX2, 68c max temp.

15c coolant temp.

 

[xkm1948]

6950X (10C/20T), all core 4.0GHz and single core 4.2GHz
Air cooling
Multi: 12438
Single: 1053

 

[heky]

My results (room temp 20-22°C):

• ST: 1314 (50°C)
• MT: 19896 (66°C)
• ST and MT test ran for 10 minutes

Setup:

• Ryzen 9 3900X, CCD0 @ 4.4Ghz CCD1 @ 4.2Ghz @ 1.2625v
• Patriot Viper Steel 16-16-16-32-48 1T 4x8GB, FCLK 1900
• Windows 10 20H2 build 19042.630
• AGESA 1.1.0.0 Patch C, BIOS Ver. 7C36v1B (MSI X570 Prestige Creation), AMD Chipset Drivers 2.10.13.408, 1usmus Ryzen Universal power plan
• EKWB Phoenix 360 with fans @ 50%, Pump @ 100%
• BIOS Settings:
  - CPPC: Enabled
  - CPPC Preferred Cores: Enabled
  - Global C-State Control: Enabled
  - Power Supply Idle Control: Low Current Idle
  - DRAM Voltage: 1.38v
  - D.O.C.P. Profile: Disabled (4400-19-19-19-39)
• 

 

[fero90]

Setup:

• Ryzen 5 5600x Stock PBO Auto
• Corsair LPX 3600 mhz 16-19-19-39 1:1
• Windows 10 2004
• MSI B550M Mortar AGESA 1.1.0.0
• AIO CM 120mm

 

[Sybaris_Caesar]

Ryzen 5 3600
Max PBO
Cryorig H7 Quad Lumi

R23 ST

R23 MT

R20

 

[HTC]

Seems scaling past 128threads still isn't great, 66k if I run 128threads, 88k at 256threads.
still better than r20 I suspect as I could get nowhere close to Overclocked 3990x.

dual Epyc 7742 (unlocked), overclocked to 3.02ghz all core on air. 3200mhz ecc.

View attachment 175376

You built up a "small lead" there ...

 

[frostybe3r]

Lmao, it's quite funny how biased this benchmark is for AMD, 4.3GHz 3970X gets 50K, 4.5GHz W-3175X 40K.

In the real world(blender, vray, premier pro, etc) w-3175x beats the Threadripper...

 

[Stewen1967]

ROG STRIX Z490-F GAMING
10700K (Stock) (Idle 27 Max 68 Celsius) No OC !
Predator 3200 Mhz (16 cl) 4 X 8 (32 GB) XMP
Thermaltake water 3.0 360 ARGB Sync
C-700P

 

[rethcirE]

How is it my i7-8750H @ 4.1GHz ST is scoring slightly higher than a 6950X @ 4.2GHz ST? 1053 vs 1105 seems beyond margin of error?

The MT scores are more expected; 12438 (10C/20T) vs 8164 (6C/12T). Somewhat explained also by 140W vs 45W TDP. Just unsure how the single thread test is being judged.

 

[tabascosauz]

How is it my i7-8750H @ 4.1GHz ST is scoring slightly higher than a 6950X @ 4.2GHz ST? 1053 vs 1105 seems beyond margin of error?

The MT scores are more expected; 12438 (10C/20T) vs 8164 (6C/12T). Somewhat explained also by 140W vs 45W TDP. Just unsure how the single thread test is being judged.

8750H is Coffee Lake (Skylake). 6950X is Broadwell-E, Skylake had a number of IPC improvements over BW-E. Now that scores are inflated, differences semm bigger than before.

No telling how much stuff he had on in the background as well.

 

[tvamos]

Ryzen 5 3600
Max PBO
Cryorig H7 Quad Lumi

R23 ST
View attachment 175408
R23 MT
View attachment 175409


R20
View attachment 175410

That's a bit low on multi R23, mine got 9080, and on computerbase de there are a lot bigger scores, 9500+

 

[mirrormax]

Lmao, it's quite funny how biased this benchmark is for AMD, 4.3GHz 3970X gets 50K, 4.5GHz W-3175X 40K.

In the real world(blender, vray, premier pro, etc) w-3175x beats the Threadripper...

Does it though? Most benchmarks I've seen 3970 either beats it or is trading blows at much lower cost and tdp. I could be wrong though but I'd like to see the benchmarks

 

[frostybe3r]

Does it though? Most benchmarks I've seen 3970 either beats it or is trading blows at much lower cost and tdp. I could be wrong though but I'd like to see the benchmarks

All the reviews you've watched are of people being incapable of using the processor properly...

Linus: takes 1.5 hours to rebuild a prebuilt, loses two memory channels and runs it at 2.3ghz, 800mhz below stock

der8aeur: runs at 4.4ghz(not even intensive avx2) x24 cache 105c with ridiculously high latency ram

toms hardware: runs high latency ram without filling all the ram slots on the t-topology board, which in itself diminishes performance, runs at 110c 3.6ghz avx2, basically useless for a review

Jayztwocents: 6 dimms in a 12 dimm t top board and hits 100c+, didn't even test anything meaningful

Gamersnexus: ran at 110c and throttled in blender rendering the results incorrect for the clocks shown, ran crappy ram too

Ihad a 3970x in the past and sold it for the Xeon, at the same clocks the w3175x beats the Threadripper in rendering applications, avx and memory intensive workloads, I mean, you can challenge my own use of the processor if you want but I've already done it time and time again, zen 2 doesn't even have 256 data paths, 256 bit registers either, it's not a complete avx256 instruction, plus I use avx512 quite alot myself for rendering, simulation and transcoding.

Reviewers had absolutely no idea what they were doing with the system and it's genuinely annoying as someone who uses it daily.

3970X is a nice cheap alternative for HEDT 3d rendering/video, the 3990x is sortve a meme with the quad channel and high latency design, not for me, uh.

Also I run this xeon at 4.5ghz avx2 no offset, it's pretty unmatched for performance in real world application

 

[Tomgang]

One last bechmark for a decade old CPU now, before it will be taken out of duty as my main system. 2021 will be Zen 3 for me.

I7 980X oc to 4.4 GHz all core. AIR cooled by Noctua NH-D14.

 

[T4C Fantasy]

One last bechmark for a decade old CPU now, before it will be taken out of duty as my main system. 2021 will be Zen 3 for me.

I7 980X oc to 4.4 GHz all core.

you didnt state cooling

 

[Tomgang]

you didnt state cooling

Cooling is AIR (Noctua NH-D14).

 

[T4C Fantasy]

Why isn't R23 a separate thread?

because ease of access to all scores is more convenient for all of you, and me.

I like the new 10 minute test as it takes thermal throttling to account. I think I need a new case and cooler.View attachment 175389

I boosted my pump and fans and did another run. Currently just running two b-die sticks. Let's see the difference when my new memory arrives, when actually using quad channel memory.View attachment 175395

i will add this strictly because these CPUs are rare to see in this thread, but i need to know your cooling and a cpuz shot anyways and clocks

 

[Sybaris_Caesar]

That's a bit low on multi R23, mine got 9080, and on computerbase de there are a lot bigger scores, 9500+

Mine's an early sample. Maybe a few weeks after release. So it doesn't clock very high. I saw 3975 Mhz on few cores during MT load, while 3925 Mhz was median I would say.
And there are some 8600 point score entries on computerbase.de listing as well: https://www.computerbase.de/2020-11/cinebench-r23-community-benchmarks/

 

[agent_x007]

Xeon E5 1680 v2 @ 4,2GHz (Air cooling).
Result : 7995 MT/804 ST
(tested with default or 10 min stress setting).

 

[Sybaris_Caesar]

Btw @T4C Fantasy, my post also includes Cinebench R20 scores. Can you add it to the R20 table?

 

[T4C Fantasy]

Btw @T4C Fantasy, my post also includes Cinebench R20 scores. Can you add it to the R20 table?

im limiting what is added, only R23

people will see it when they click on your R23 score though, which is one of the reasons i add links

 

[mirrormax]

All the reviews you've watched are of people being incapable of using the processor properly...

Linus: takes 1.5 hours to rebuild a prebuilt, loses two memory channels and runs it at 2.3ghz, 800mhz below stock

der8aeur: runs at 4.4ghz(not even intensive avx2) x24 cache 105c with ridiculously high latency ram

toms hardware: runs high latency ram without filling all the ram slots on the t-topology board, which in itself diminishes performance, runs at 110c 3.6ghz avx2, basically useless for a review

Jayztwocents: 6 dimms in a 12 dimm t top board and hits 100c+, didn't even test anything meaningful

Gamersnexus: ran at 110c and throttled in blender rendering the results incorrect for the clocks shown, ran crappy ram too

Ihad a 3970x in the past and sold it for the Xeon, at the same clocks the w3175x beats the Threadripper in rendering applications, avx and memory intensive workloads, I mean, you can challenge my own use of the processor if you want but I've already done it time and time again, zen 2 doesn't even have 256 data paths, 256 bit registers either, it's not a complete avx256 instruction, plus I use avx512 quite alot myself for rendering, simulation and transcoding.

Reviewers had absolutely no idea what they were doing with the system and it's genuinely annoying as someone who uses it daily.

3970X is a nice cheap alternative for HEDT 3d rendering/video, the 3990x is sortve a meme with the quad channel and high latency design, not for me, uh.

Also I run this xeon at 4.5ghz avx2 no offset, it's pretty unmatched for performance in real world application

ok thats interesting, so all the reviewers struggled with overheating? i noticed you run a chiller of some sort? how does it do on air?
do you have some more benchmarks of your system somewhere? vray/corona/pugets premier etc.

 

[biffzinker]

Cooling: Cooler Master Hyper 212 Black Edition
Clockspeed: 4,275 MHz 1.24V (ClockTuner)

 

[mirrormax]

forgot cpu-z also gained abit more score ignore cpuz thinking its a TR and the clocks when OCed dont display right. they are 7742s just unlocked oems OCd to 3ghz with noctua air coolers.

 

[Det0x]

damn 15° ambient ...

At least i'm honest

@ T4C Fantasy

I have a few r20 scores in post #1801 which can be added