****Blender Ryzen render battle benchmark****

[FlanK3r]

What do you need? First the render software Blender in version 2.78a
http://mirror.cs.umn.edu/blender.org/release/Blender2.78/blender-2.78a-windows64.msi

And after render scene:
https://ulozto.cz/!v5ELMDZKE3k6/ryzengraphic-27-blend this scene is with 200 samples predefined
AMD official render scene with 150 samples

In default is render test set to 200 samples quality. AMD útested the CPU Ryzen ES with 150 samples and also at 100. AMD RyZen 3.4 GHz 8C/16T got the scores

100 samples = 25s
150 samples = 36s


Show your results+info about samples and CPU-Z. Enjoy guys

 

[Raelgun]

5820 K @ 4.3 Ghz

100 samples = 27.69
150 samples = 40.40
200 samples ( default) = 53.53

No other settings changed, only changed the sample number.

 

[Jeps]

4770K @ 4.3Ghz

100 samples = 40.77
150 samples = 60.15

Not too bad

 

[Tee Hee Johnson]

Can probably jump into the Zen thread and get lots of benchmarks already done

 

[jagga1968]

plenty of results here

https://www.reddit.com/r/Amd/comments/5i6w0w/submit_your_blender_test_for_comparison/

btw amd uploaded the wrong file.

 

[jagga1968]

oops

 

[joeyjojo]

5820 K @ 4.3 Ghz

100 samples = 27.69
150 samples = 40.40

So 11-12% faster than your 5820K, but running 20% slower clock speed with 33% more cores...

In terms of cores*clock speed advantage that works out as the Ryzen having 5% more, so the 6-7% gap must be architecture improvements. Not bad, but not amazing considering that's versus Haswell. Probably a dead heat against Skylake and newer.

Edit: were you running 6-way or 12-way? Isn't Ryzen supposed to be SMT as well? If you both had hyperthreading on it's a better comparison.

 

[Fireblade2K4]

AMD FX-8350 @ 4.5Ghz

100 samples = 1m20s
150 samples = 1m55s
200 samples = 2m33s

 

[Mrjobby]

i7 920 @ Stock

100 1:30
150 2:16

Still got it lol

 

[Sixense]

6700K @ 4.4

100 - 35.31
150 - 52.54

Looks fine

 

[humbug]

AMD messed up, they uploaded the file with Default settings which is 200 Samples, the Ryzen benchmark was actually done at 150 Samples.

AMD uploaded the correct file here http://download.amd.com/demo/RyzenGraphic_27.blend

My result...

I have never compared my FX-8 CPU's to the 4690K but in terms of performance i didn't notice any diffrence...

Anyway.... my 4690K @ 4.5Ghz / 4.2Ghz UnCore / 2400Mhz 16GB CL11

Edit 150 samples, thats what AMD ran.

1:26.23

3.4Ghz 8 Core Zen is nearly 3x faster, impressive.

 

[psychas]

100 samples = 38.51 s
150 samples = 57.71s

 

[implosiveturnip]

I'll be giving this a bash on wide(ish) array of intel CPUs this evening

Though looking at these results, I think I can finally justify an upgrade from my 3770k, provided the performance uplift we're seeing is ubiquitous; across a majority of games & applications and we're not just looking at a best case only scenario

 

[Deleted member 66701]

4770k @ 4.8ghz

100 = 36.52s
150 = 53.89s

Hoping Ryzen does in fact turn out to be a viable upgrade route for me. Now I'm using Photoshop, Premier, After Effects and 3Ds Max extensively, a core upgrade would be worthwhile. Plus the fact I ran out of memmory editing a 1.5gb PSD file yesterday so I need more memory as well!

 

[Yaayuh!]

What we really need are some 8/16 users to bench, it's apple to oranges at the moment. Or in fact, just single threaded benchmarks.

 

[humbug]

What we really need are some 8/16 users to bench, it's apple to oranges at the moment. Or in fact, just single threaded benchmarks.

Why single threaded?

Anyway..... Kaap? you have one right?

 

[chris182008]

e5-2680 (ES) x 2. 3.1GHz all core turbo. For 8c/16t I limited blender to 16 threads.

100 samples
8c/16t 30.08
16c/32t 18.07

150 samples
8c/16t 46.9
16c/32t 26.82

 

[fastboy]

1:29.70 i5 6400 oc to 4.3ghz for 150 samples,that sound about right?

 

[tyler_jrb]

5930k @ 4.7ghz

150 Samples = 37.9s

Not bad for my 6c/12t CPU.

 

[whyscotty]

[email protected]

150 samples = 31.52 secs

Spoiler