Maxon CinebenchR23

[Leeghoofd]

K gents Maxon just released a new Cinebench version23

By default it is a 10 minute stress test looping the rendered scene similar to R20

In the advanced tab we can turn off the time duration so it renders just one benchmark like the previous Cinebench versions. Or we go half way and choose a custom duration of eg  5minutes

 

 

[damric]

Nevermind. Change my vote from 10 minutes to a single pass through. No time for that.

[Jumper118]

Don't see much point adding new benchmarks when pretty much all of them are in beta already and don't give points. Haven't even merged benchmate and normal rankings. 

[gt403cyl2]

I would like to see bugs fixed and things cleaned up before adding new benches personally.
That said if it's going to be added, Default gets my vote.

[Rauf]

I voted not another cinebench, but I guess if it is added a fixed time would be interesting. But maybe a  shorter time. Sure 3D benchmarks are a few minutes long but at least you can watch the "game" and not just stare at the termometer

 

Also I guess which instructions are used matters. AVX? As I understand it AVX is still kind of a cpu killer.

Edited November 11, 2020 by Rauf

[max1024]

Something NEW and especially Cinebench it is always Good ?  On my opinion, One run will be enough. Overclocking and stability are two different substances. I've voted for "Add Cinebench R23 and turn off minimum test duration"

[Leeghoofd]

1 hour ago, Jumper118 said:

Don't see much point adding new benchmarks when pretty much all of them are in beta already and don't give points. Haven't even merged benchmate and normal rankings. 

We are working on that, tomorrow the programmer has a meeting with Frederik how to move them Beta's to released state and how to introduce new benchmarks (like Port Royal, Geek5 and maybe this one)

Points will be attributed for most benches in the new HWBot version

[Leeghoofd]

1 hour ago, gt403cyl2 said:

I would like to see bugs fixed and things cleaned up before adding new benches personally.
That said if it's going to be added, Default gets my vote.

So you will have to be very patient before everything will get properly sorted before we introduce anything new ?

[Doomed83]

I voted to implement without minimum duration, but would also like Benchmate to be mandatory provided it's compatible.

[jab383]

If it's rendering the same scene as R20 and testing the same processor features, what is added by running this bench?  If it's really testing things not included in the other versions of Cinebench, then I think one shot rendering is best.

 

[GeorgeStorm]

Please not another cinebench.

[Leeghoofd]

Well Matt analysed already the R23 program in depth.

Render seems very very similar and the bench ticks the same way as previous versions. So for AMD and older Intels BenchMate remains mandatory.

Now if the render and CPU load is identical I don't see the point to run it with a single render. As it will be the same as R20 but with a different score output. I don't get it why Maxon did not opt to go for another render.

 

[_mat_]

To get the ball rolling here as well:

While analyzing CINEBENCH R23 I had an idea. It seems like none of us want to embrace a new benchmark that brings nearly the same workload and result image as its predecessor. So why don't we customize a workload to our needs?

Ladies and gentlemen, this could be the CINEBENCH R23 - HWBOT Edition:

1. Let's make the workload HWBOT themed. There are these great renderings from somebody here for various competitions!
2. We can roughly define the length of the benchmark, its memory consumption (for better tweaking with MAXMEM) and its heat production
3. We let BenchMate secure it for us, so that nobody can touch the scene (this is very much necessary for all CINEBENCH versions btw)

What you see in the picture: The first run was with the original CB R20. It took about 1 minute and 5 seconds, needed 2 GB of process memory and got the CPU to 73 °C. Second run was with a demo scene I prepared to work with CINEBENCH. It took exactly 2 minutes, only needed 1 GB and reached 76 °C in the end. This is on a 10700K - 8C/16T @ 4,7G.

Let me know what you think!

My FB post: https://www.facebook.com/matthias.zronek/posts/10157528951688199
Demo scene credits: https://www.thepixellab.net/freebies

[flanker]

It is not bad test, but...R11, R15, R20 (but we need points for R20 results I think)

Good one for R23 is it loop test-good for partial stability and it could be interesting for editors. Because with defaults specs of CPUs can be interesting how Intel CPUs drop the frequency after 10 minutes and average results fater 10minutes can be much lower than on one run (compared to Ryzens with more consistent frequencies in long term bases.

 

Maybe  add to database for fun, without points now

[_mat_]

Benchmarks or at least associated points should change more often anyways. Yearly basis for example, so even the pros have to step up their game once in a while.

I also think it's a good idea to move on from R20 directly to R23. Would only make sense with the custom R23 version though.

[Jumper118]

17 hours ago, Leeghoofd said:

We are working on that, tomorrow the programmer has a meeting with Frederik how to move them Beta's to released state and how to introduce new benchmarks (like Port Royal, Geek5 and maybe this one)

Points will be attributed for most benches in the new HWBot version

That is great news. 

[Splave]

make y-cruncher great again, points in 2021

[chispy]

14 hours ago, _mat_ said:

To get the ball rolling here as well:

While analyzing CINEBENCH R23 I had an idea. It seems like none of us want to embrace a new benchmark that brings nearly the same workload and result image as its predecessor. So why don't we customize a workload to our needs?

Ladies and gentlemen, this could be the CINEBENCH R23 - HWBOT Edition:

1. Let's make the workload HWBOT themed. There are these great renderings from somebody here for various competitions!
2. We can roughly define the length of the benchmark, its memory consumption (for better tweaking with MAXMEM) and its heat production
3. We let BenchMate secure it for us, so that nobody can touch the scene (this is very much necessary for all CINEBENCH versions btw)

What you see in the picture: The first run was with the original CB R20. It took about 1 minute and 5 seconds, needed 2 GB of process memory and got the CPU to 73 °C. Second run was with a demo scene I prepared to work with CINEBENCH. It took exactly 2 minutes, only needed 1 GB and reached 76 °C in the end. This is on a 10700K - 8C/16T @ 4,7G.

Let me know what you think!

My FB post: https://www.facebook.com/matthias.zronek/posts/10157528951688199
Demo scene credits: https://www.thepixellab.net/freebies

Excellent my friend Matt , " this is the way " ( In mandalorian voice ? )

I like it , this idea of yours is great !

[chispy]

@Leeghoofd , today appears another free contender to add to hwbot  V-Ray 5 can be used for cpu and gpu benchmarks.  What you guys think ?

https://www.tweaktown.com/news/76177/test-your-geforce-rtx-series-graphic-card-with-ray-5-benchmark/index.html

[Leeghoofd]

39 minutes ago, Splave said:

make y-cruncher great again, points in 2021

 

32 minutes ago, chispy said:

@Leeghoofd , today appears another free contender to add to hwbot  V-Ray 5 can be used for cpu and gpu benchmarks.  What you guys think ?

https://www.tweaktown.com/news/76177/test-your-geforce-rtx-series-graphic-card-with-ray-5-benchmark/index.html

Thread derailers alert !!!

[Splave]

This one maybe bit more interesting? 

 

[unityofsaints]

On 11/11/2020 at 8:27 PM, Leeghoofd said:

We are working on that, tomorrow the programmer has a meeting with Frederik 

[_mat_]

14 hours ago, Splave said:

This one maybe bit more interesting? 

 

To clarify and make myself look bad again and again, over and over: You can easily find a way to circumvent the current screenshot rule. 

The only technique mitigating this vulnerability is the validation of all files that are used for the workload. There is no other way. It also avoids real renderings of the workload with slightly decreased image quality. For CB R11.5 and R15 you can edit and put your own logic in there, basically just setting the result value you want. This also goes away if file integrity is always checked.

Sorry to bring the bad news. There are more vulnerabilities btw, but I leave it at that so other benchmarks are not harmed.

[Matsglobetrotter]

I consider CB R23 should be included, minimum length. but get rid of R20 and the hack version as we never got points for either. I assume AMD will be on 100 cores soon ? I enjoy these more than 3 versions of geekbench.

I wish someone could convince UL/Futuremark to Change their PCMArk10 to something that has a overclockers run.. I bought it and think used it 10 times as its way to long and boring but like the concepts of PCMark 7 and PCVantage (even if vantage is a drag of 20 minutes) 

[FatBoyNotSoSlim]

On 11/13/2020 at 2:05 AM, Splave said:

make y-cruncher great again, points in 2021

First you had my respect, now you have my attention.

MYCGA 2021.