Friday, September 2nd 2022

AMD Ryzen 9 7950X Tested in Cinebench R23

A Cinebench R23 picture of AMD's recently announce Ryzen 9 7950X CPU having been put through its paces have appeared online via a post on Baidu, which has been taken down since the picture was posted. However, courtesy of @harukaze5719 it lives on, on Twitter and gives us a first glimpse into the Cinebench R23 performance of the upcoming CPU. The CPU is said to have been air cooled, so it's possible that we'll see even higher benchmark numbers with better cooling, so take these numbers with a pinch of sodium chloride, just to be on the safe side. The test system was also using Windows 10, so there's the potential of some extra performance by changing to Windows 11 here as well.

In the single score test, the Ryzen 9 7950X scores 2,205 points, which is in line with Cinebench R23 leaks for Intel's upcoming Core i9-13900K CPU, if a smidgen slower. The multi-core score is obviously not going to compete with Intel's Core i9-13900K due to the overall lower core count, but at 29,649, but it's ahead of the Core i9-12900K by a decent margin. It'll be interesting to see how AMD positions the 7000-series of CPUs, as although it seems like the company has done a good job in improving the overall performance compared to the 5000-series, it's not quite enough to take the performance crown this time around, if these early benchmark leaks from both sides are anything to go by.

Update 10:27 UTC: A new picture hjas appeared where the CPU has been kitted out with better cooling at the multi-core score has jumped from 29,649 to 36,256, which makes it competitive with the Core i9-13900K scores that have leaked in the past.
Sources: @harukaze5719, @henry41224
Add your own comment

83 Comments on AMD Ryzen 9 7950X Tested in Cinebench R23

#26
HTC
ice_vThe leaker also posted this:
Both appear to be real and the reason for the low performance is explained in this anandtech forum post, in the edits part.
rainzorThe score is crap because the CPU throttled under air cooler. With 360 AIO it scored just over 36K.

That is supposed to be the exact same CPU under better cooling.
Posted on Reply
#27
ModEl4
Too low to be real, it corelates with the score that 7950X would have if it had active only 12 from the total 16 cores.(essentially close to the 7900X score)
Posted on Reply
#28
ARF
Err, that means the decision to increase the TDP from 105 to as high as 170 watts and beyond is a bad decision and will cause all kinds of problems.

The Zen 4 is the new Bulldozer?
Posted on Reply
#29
TheLostSwede
News Editor
HTCBoth appear to be real and the reason for the low performance is explained in this anandtech forum post, in the edits part.


That is supposed to be the exact same CPU under better cooling.
I've added an update to reflect this.
Posted on Reply
#30
demian_vi
ARFErr, that means the decision to increase the TDP from 105 to as high as 170 watts and beyond is a bad decision and will cause all kinds of problems.

The Zen 4 is the new Bulldozer?
no it doesn't :laugh::laugh:

easy with the hyperboles
Posted on Reply
#31
Unregistered
napataDid Intel ever say what the E(efficiency) refers too? I think it refers to efficency in area as they are unmatched in that while P-cores are seemingly more energy efficient.
My understanding is they are power efficient as well as not taking too much space.
Efficient-cores are:
  • Physically smaller, with multiple E-cores fitting into the physical space occupied by one P-core.
  • Optimized for multi-core performance-per-watt—delivering scalable multithread performance and efficient offload of background tasks.
  • Capable of running a single software thread.
  • Capable of 40% more performance when running at the same power as a single Skylake core4.
#32
ARF
Xex360My understanding is they are power efficient as well as not taking too much space.
Efficient-cores are:
  • Physically smaller, with multiple E-cores fitting into the physical space occupied by one P-core.
  • Optimized for multi-core performance-per-watt—delivering scalable multithread performance and efficient offload of background tasks.
  • Capable of running a single software thread.
  • Capable of 40% more performance when running at the same power as a single Skylake core4.
Is a system with big-small cores more responsive in Windows?
Posted on Reply
#33
big_glasses
big doubt that it would be so throttle between D15 and 360 Aio
to my understanding cinebench is 10min(?)

With Aida64 FPU test, there is less than 10c difference between ND-D15 and best AiO, in 10900k test ("This results in the CPU going from the 125-watt limit defined by Intel all the way up to 200-watts under long-term heavy load"). (I don't know how long Aida64 runs)
OC Blender have 71c (Aio) vs 77c (NH-D15) over 15 minutes.
www.techpowerup.com/review/g-skill-enki-360-aio-cpu-closed-loop-liquid-cooler/7.html

To me it seems that the short run of cinebench shouldn't be long enough to "overload" the cooler
---
I just realized it didn't specify the air cooler, other than being an air cooler.... that can mean everything from wraith level(or lower) to NH-D15 level
Posted on Reply
#34
HTC
big_glassesI just realized it didn't specify the air cooler, other than being an air cooler.... that can mean everything from wraith level(or lower) to NH-D15 level
Not just that: apparently, an old BIOS version was also used in the original benchmark.

HOWEVER, dunno if it was the same BIOS used in the latest benchmark or a newer one.
Posted on Reply
#35
TheoneandonlyMrK
ARFErr, that means the decision to increase the TDP from 105 to as high as 170 watts and beyond is a bad decision and will cause all kinds of problems.

The Zen 4 is the new Bulldozer?
ARFIs a system with big-small cores more responsive in Windows?
I didn't have you down as a troll before, was I wrong, do you need ignoring?!

Bad day perhaps.
Posted on Reply
#36
ARF
TheoneandonlyMrKI didn't have you down as a troll before, was I wrong, do you need ignoring?!

Bad day perhaps.
Good luck!
Waiting for the miracle reviews :D
Posted on Reply
#37
gagegfg
stimpy88My 5950X gets a higher multi-core score than that. I also get 60GBs memory bandwidth. So it's either fake, or setup by a fool.
that cpu is overclocked
Posted on Reply
#38
DeathtoGnomes
So all the weak suckers fell into this leak/speculation trap. Some people need to reminded to wait for reviews, some dont. :banghead:
Posted on Reply
#39
gagegfg
Xex360Two cores are after than one core with 2 threads.
Plus E cores are not all that powerless, I believe 12th gen E cores were similar to Kaby Lake, now with better node and other enhancemens I think they will be even more potent.
Unfortunately they don't deserve their E prefix, as Intel gets trashed by AMD for battery life.
the E core are for efficiency. The P cores are high power and with HT and higher frequency they could outperform two E cores.
Posted on Reply
#40
ARF
DeathtoGnomesSo all the weak suckers fell into this leak/speculation trap. Some people need to reminded to wait for reviews, some dont. :banghead:
What if the reviews actually paint an even worse picture?

I told you that AMD will not look good, there is no reason to expect any good results, especially given the specifications of the new SKUs. How much can a poor 6-core be? In the best case it will trade blows with the old Ryzen 7 5800X.

Especially when you so aggressively shout wait for the reviews.

One month to wait, though.
Posted on Reply
#41
usiname
TheoneandonlyMrKI didn't have you down as a troll before, was I wrong, do you need ignoring?!

Bad day perhaps.
Lol, this is the other troll "HenrySomeone", why do you think he stopped shitting under the AMD articles?

By the way how st you should be to beleave that 7950x will score less than 30k in R23? Even ryzen 5950x at 4.3ghz can score higher
Posted on Reply
#42
TheoneandonlyMrK
usinameLol, this is the other troll "HenrySomeone", why do you think he stopped shitting under the AMD articles?
Oh riighht, taking turns.
ARFGood luck!
Waiting for the miracle reviews :D
It's better than pulling doom and bullshit out your arse over every leak but you do you.
Posted on Reply
#44
ixi
Bladi, bladi who cares about 7950x or 7900x... give us cpu from 50 up to 200€...
Posted on Reply
#45
Unregistered
ARFIs a system with big-small cores more responsive in Windows?
I have no idea, though it would be interesting to find a way to test this.
#46
noname00
LenneJudging a product by leaks is totally pointless if you ask me.
The only thing that worrying regarding the Zen4 lineup is the lower price of 7950x compared to 5950x. This almost never happens with a product that outperforms the competition. As for the leaks&rumors - I don't care at all.

We should wait for reviews for both AMD and Intel CPUs and actual prices in shops before drawing conclusions.
Posted on Reply
#48
A Computer Guy
LenneJudging a product by leaks is totally pointless if you ask me.
There are no leaks, only strategically placed marketing tactics between competitors.
Posted on Reply
#49
AM4isGOD
There's a proper us vs them thing going on on TPU. Never used to be like this, such a shame, humans can be such morons sometimes. For new visitors to TPU it's going to show what pleasant bunch we are indeed.

Only 2 pages and it is already looking like it will need a mod clean up.
Posted on Reply
#50
A Computer Guy
rainzorThe score is crap because the CPU throttled under air cooler. With 360 AIO it scored just over 36K.

Fast forward to 2023 some headlines I predict...
  • AMD increases industry sales of AIO's by 147% as performance chasers go wild for Zen4.
  • Are AIR coolers doomed? Noctua finally creates brown and beige AIO with matching fans.
Posted on Reply
Add your own comment
Jul 16th, 2024 02:37 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts