Blender 3.3.0 Benchmark

[SchumannFrequency]

Software: ROSA Fresh Desktop 12.4 -- LXQt -- Nvidia proprietary driver -- XFS as root
Hardware: Intel 12600KF (stock) -- Kingston 6200 MHz CL35 -- GTX 650 1GB -- BIOSTAR B760MZ-E PRO -- Antec P6 -- Xilence XP550 -- ARCTIC i35 -- EVO 850 500GB

I had just done some heavy multi-core benchmarks before doing this test, so the PC was fairly warm and did not reach its maximum potential in the test anyway.
Think of it as a result that indicates the absolute minimum values this CPU achieves in the test.

What strikes me is that the Ryzen 7600X, the only AMD competitor to the Intel 12600KF currently, does not perform as well in this test despite being a more expensive CPU:

AMD Ryzen 5 7600X CPU Review - Page 5 of 9

AMD's new Zen 4 based Ryzen 5 7600X CPU is here, and we have all the benchmarks, synthetic, rendering, and gaming with a 10 CPU comparison.

www.thefpsreview.com

You can take the test on this page. https://opendata.blender.org/
When you post, please include your specific hardware and software.

 

[Franz]

gpu:

System specs on perfil

 

[SchumannFrequency]

If I do this test when the CPU is less hot it scores slightly higher.

I also tested the performance of the Nvidia GTX 650 1GB and it scores more than 4 times lower than the Intel 12600KF.

A GPU is not necessarily faster than a CPU for Blender. It is probably only from the moment a GPU has 7.5 TFLOPS FP32 (float) processing power that it is going to be faster than any CPU.

 

[Marcus Vinicus]

System as per signature
Windows 11 Pro 23H2 22635.2915

 

[Det0x]

7950X / 8152MT/s CL32 / Win11

3090 / 7950X / 8152MT/s CL32 / Win11

 

[SchumannFrequency]

Software: Clear Linux -- GNOME -- Wayland -- nouveau GPU driver -- F2FS as root
Hardware: Intel 12600KF (stock) -- Kingston 6200 MHz CL35 -- GTX 650 1GB -- BIOSTAR B760MZ-E PRO -- Antec P6 -- Xilence XP550 -- ARCTIC i35 -- EVO 850 500GB

 

[SchumannFrequency]

Hardware: Intel 12600KF (stock) -- Kingston 6200 MHz CL36 -- Sapphire RX 7600 -- BIOSTAR B760MZ-E PRO -- Antec P6 -- Xilence XP550 -- ARCTIC i35 -- EVO 850 500GB
Software: Calculate Linux, KDE Plasma, Mesa open-source driver, XFS file system

On average, Calculate Linux seems to outperform most Linux systems in this Blender test. The differences are rather subtle.
Maybe it's just because I've slightly adjusted my RAM frequency and timings that it's a little faster now.

I now outperform the more expensive R5 7600X quite strongly in each of the three tests, when you compare it to these results:

AMD Ryzen 5 7600X CPU Review - Page 5 of 9

AMD's new Zen 4 based Ryzen 5 7600X CPU is here, and we have all the benchmarks, synthetic, rendering, and gaming with a 10 CPU comparison.

www.thefpsreview.com

 

[DaemonForce]

Seems legit. :/

 

[SchumannFrequency]

View attachment 343182

Seems legit. :/

You tested the wrong version.
The idea is for everyone to test version 3.3.0 so that the results are comparable.

This is my result when I use the CLI version of the benchmark.

 

[DaemonForce]

Nevermind I see it

I guess there's some difference. Is it enough to matter? We'll see.

 

[SchumannFrequency]

Nevermind I see it

From the name of the .tar.gz/zip/7z/rar file it seems that it only tests version 3.1.0.
But if you run the script you can choose the version:

I guess there's some difference.

You score much higher now. Because you used the correct version.

 

[AVATARAT]

7900x

 

[Arctucas]

 

[jiajo]

CPU

GPU

 

[Ivan Bliminse]

 

[Ikkyusan]

Intel Core i5-13500 with Windows 11
https://opendata.blender.org/benchmarks/6846fbb5-946b-4839-a695-d3cc6b8e7a48/

 

[SchumannFrequency]

Intel Core i5-13500 with Windows 11
https://opendata.blender.org/benchmarks/6846fbb5-946b-4839-a695-d3cc6b8e7a48/
View attachment 347932

The Intel Core i5-13500 is 7% faster in multi-core than the i5-12600KF.
And the i5-12600KF is 5% faster in single-core performance.

I suspect that they perform very similarly on average with standard use.

 

[Ikkyusan]

I've bench again in 2024, with RAM tuning by DDR5-5400 CL34-39-39-78 & tREFI = 65535
(In previous bench: DDR5-5200 CL40-40-40-80 & tREFI = Auto)

Now get over 300 points

https://opendata.blender.org/benchmarks/88dcf13e-b396-47e0-8f5e-79b6430e4970/

 

[SchumannFrequency]

I've bench again in 2024, with RAM tuning by DDR5-5400 CL34-39-39-78 & tREFI = 65535
(In previous bench: DDR5-5200 CL40-40-40-80 & tREFI = Auto)

Now get over 300 points

https://opendata.blender.org/benchmarks/88dcf13e-b396-47e0-8f5e-79b6430e4970/
View attachment 348009

Suppose you used RAM speed >8000 MHz, your result would probably also increase by a few percent.

There is another Blender thread that I made and I am curious whether you can achieve a better result here than my Intel 12600KF under Linux.

Blender 4.0.1 BMW27 demo test

Render time: 119.2 seconds This test is famous in part because Intel or AMD have frequently used the test to demonstrate the advances of their latest-generation CPUs. Hardware: Intel 12600KF (stock) -- Kingston 6200 MHz CL36 -- Sapphire RX 7600 -- BIOSTAR B760MZ-E PRO -- Antec P6 -- Xilence...

www.techpowerup.com

 

[Arctucas]

GPU

 

[SchumannFrequency]

Hardware: Intel 12700KF (stock) -- G.SKILL RIPJAWS @3600 CL18 (stock) -- Sapphire RX 7600 -- ASRock B760M-ITX/D4 WiFi -- fractal design DEFINE NANO S -- bequiet! SYSTEM POWER 10 550W -- DeepCool AG500BK ARGB -- EVO 850 500GB
Software: Calculate Linux, PeKWM, Mesa open-source driver, XFS file system

monster: 181.678 samples per minute
junkshop: 94.124 samples per minute
classroom: 71.878 samples per minute

I haven't tuned all the BIOS settings for maximum performance yet.

 

[SchumannFrequency]

Hardware: Intel 12700KF (stock) -- G.SKILL RIPJAWS @3600 CL18 (stock) -- Sapphire RX 7600 -- ASRock B760M-ITX/D4 WiFi -- fractal design DEFINE NANO S -- bequiet! SYSTEM POWER 10 550W -- DeepCool AG500BK ARGB -- EVO 850 500GB
Software: Artix Linux, LXQt, Mesa open-source driver, F2FS file system

monster: 184.915 samples per minute
junkshop: 100.135 samples per minute
classroom: 73.251 samples per minute

 

[Ferrum Master]

I did few weeks ago. Took from online submitted result.

samples_per_minute: 237.54118723269517,
samples_per_minute: 159.8607014634951,
samples_per_minute: 116.26322483003287,

And GPU.

• samples_per_minute: 2123.5943215660523
• samples_per_minute: 1165.0576138559823
• samples_per_minute: 996.0193849059312

 

[nuninho]

Running W11 22H2, CPU@stock with E cores disabled and GPU non-OC version@stock speeds with 546.17WHQL:

Running W11 23H2, CPU@5.2 with HT off + E cores off and GPU non-OC version@stock speeds with 555.99WHQL:

Running W10 22H2, CPU@5.2 with HT off + E cores off and GPU non-OC version@stock speeds with 556.12WHQL:

This GPU has benchmarked only, sorry. Sorry for "new" v4.1.0 but the difference of performance between v3.3.0 and v4.1.0 is none.

 

[DR4G00N]

3060 Laptop, Ver. 3.3.0, 556.12

WX 9100, Ver. 3.3.0, Pro 24.Q1

WX 9100, Ver. 4.2.0, Pro 24.Q1