Holomark 2 Released!

UPS! I didn’t notice that… I was surprised by the big results… And yes, it was not get the red graph. Are this “red graph” critical for the results?

In fact, I have never pay attention to that red graph. And I don’t know what could happen?

UPS… I have made other test, and the results, are a little diferent. But now It shows the RED GRAPH

Before I was testing with everythink turn off. KASPERSKY, and other simple apps… In this new test, I keep all thinks on: Antivirus, Winamp, and other windows apps.

I notice now, the CPU was a little more slow… I think it is because of that. HAVE A LOOK PLEASE

I think the big diference between a good GTX and a QUADRO, specially the K5000, are in the mesh tests… They handle a lot more units.

Sapphire Radeon 7870

MACBOOK PRO 15" retina

Holomark 2 v2,1,02 Beta RC 2 - Expires: 2014-5-30

Total Score: 23948
Total Runtime: 137.91 sec

GPU scores: 10190
GPU_01 - 182.60 fps - Cube 4 tests
GPU_02 - 37.50 fps - UDT Shape
GPU_03 - 61.30 fps - Wireframe
GPU_04 - 28.70 fps - Shaded
GPU_05 - 27.00 fps - Rendered
GPU_06 - 26.50 fps - Block Rendered
GPU_07 - 18 units Nurbs @ 5 fps in Wireframe
GPU_08 - 8 units Nurbs @ 5 fps in Shaded
GPU_09 - 8 units Nurbs @ 5 fps in RenderSpeed
GPU_10 - 49.50 fps - RenderMesh Render
GPU_11 - 31.50 fps - RenderMesh RenderSpeed
GPU_12 - 35.60 fps - JoinedMesh Render
GPU_13 - 28.80 fps - JoinedMesh RenderSpeed
GPU_14 - 2 units mesh @ 15 fps in Shaded
GPU_15 - 3 units mesh @ 15 fps in Render
GPU_16 - 3 units mesh @ 15 fps in RenderSpeed
GPU_17 - 27.70 fps - mesh in Rendered Studio
GPU_18 - 9.60 fps - Nurbs in Rendered Studio
GPU_19 - 24.70 fps - Block Illustration
GPU_20 - 85.50 fps - 2D single
GPU_21 - 7.20 fps - 2D massive (20x)

CPU scores: 13758
CPU_01 - 9.76 sec - Booleans and Contours
CPU_02 - 1.34 sec - Twist and Taper (UDT)
CPU_03 - 4.42 sec - Meshing Mini
CPU_04 - 0.02 sec - Extract Render Mesh
CPU_05 - 0.05 sec - Join Render Mesh
CPU_06 - 15.23 sec - Reduce Mesh
CPU_07 - 2.02 sec - Calculating Technical display
CPU_08 - 3.50 sec - Making Silhouettes

Apple Inc.
MacBookPro11,3

NVIDIA GeForce GT 750M - 2048.0 MB
DriverVersion: 9.18.13.3523

Intel® Core™ i7-4850HQ CPU @ 2.30GHz
NumberOfCores: 4 NumberOfLogicalProcessors: 8
MaxClockSpeed: 2.3 GHz

TotalPhysicalMemory: 16.0 GB

Microsoft Windows 8.1 Pro

  • None - 64-bit

Rhino 5 sr 8 64 bit

amd R9 290X
Not really good for rhino.

Own build, gigabyte motherboard /w AMD A10-6800K, 1 AMD R9 270x, 1 NVidia Geforce GTX 760, 5 monitors hooked up.

Holomark 2 v2,1,02 Beta RC 2 - Expires: 2014-5-30

Total Score: 25016
Total Runtime: 196.91 sec

GPU scores: 14960
GPU_01 - 141.80 fps - Cube 4 tests
GPU_02 - 24.20 fps - UDT Shape
GPU_03 - 39.40 fps - Wireframe
GPU_04 - 28.40 fps - Shaded
GPU_05 - 22.40 fps - Rendered
GPU_06 - 21.70 fps - Block Rendered
GPU_07 - 11 units Nurbs @ 5 fps in Wireframe
GPU_08 - 7 units Nurbs @ 5 fps in Shaded
GPU_09 - 22 units Nurbs @ 5 fps in RenderSpeed
GPU_10 - 51.80 fps - RenderMesh Render
GPU_11 - 126.60 fps - RenderMesh RenderSpeed
GPU_12 - 38.20 fps - JoinedMesh Render
GPU_13 - 188.70 fps - JoinedMesh RenderSpeed
GPU_14 - 3 units mesh @ 15 fps in Shaded
GPU_15 - 4 units mesh @ 15 fps in Render
GPU_16 - 136 units mesh @ 15 fps in RenderSpeed
GPU_17 - 43.90 fps - mesh in Rendered Studio
GPU_18 - 7.50 fps - Nurbs in Rendered Studio
GPU_19 - 17.40 fps - Block Illustration
GPU_20 - 51.50 fps - 2D single
GPU_21 - 6.50 fps - 2D massive (20x)

CPU scores: 10056
CPU_01 - 13.62 sec - Booleans and Contours
CPU_02 - 1.89 sec - Twist and Taper (UDT)
CPU_03 - 6.67 sec - Meshing Mini
CPU_04 - 0.04 sec - Extract Render Mesh
CPU_05 - 0.11 sec - Join Render Mesh
CPU_06 - 19.90 sec - Reduce Mesh
CPU_07 - 2.59 sec - Calculating Technical display
CPU_08 - 4.90 sec - Making Silhouettes

Gigabyte Technology Co., Ltd.
To be filled by O.E.M.

AMD Radeon R9 200 Series - 2048.0 MB
DriverVersion: 13.350.1005.0
NVIDIA GeForce GTX 760 - 2048.0 MB
DriverVersion: 9.18.13.3750
AMD Radeon HD 8670D - 256.0 MB
DriverVersion: 13.350.1005.0

AMD A10-6800K APU with Radeon™ HD Graphics
NumberOfCores: 2 NumberOfLogicalProcessors: 4
MaxClockSpeed: 4.1 GHz

TotalPhysicalMemory: 16.0 GB
Microsoft Windows 7 Ultimate

  • Service Pack

I suppose the GPU tests are running on my R9 270x, since it is the card with primary monitor attached to it. I wonder how to ensure tests running on the other GPUs with minimum of hassle.

iMac 27" 2010 8xAA

iMac 27" 2010 noAA

NEW Mac Pro 2013 FireGL Driver

NEW Mac Pro 2013 Catalyst Driver

It seems like the catalyst drivers makes the CPU slower on the UDT score.
@stevebaer do you have any idea why? Extract render mesh is also slower, and I find that odd too. Also making Silhouettes…

Hi @Holo!
That was a quick reply :wink:

My two cents are: I don’t think it affects CPU performance per se, I believe it’s rather that your tests cannot completely isolate CPU performance from GPU performance, they seem to alway influence each other.

I have written a few words about my Mac Pro experience here. For real life performance I would say that mesh values have less impact than nurbs values (in Holomark). Would it be an option to have Holomark only run on nurbs tests? I’ve read up on everything regarding gfx hardware and Rhino viewport performance but I’m still not sure which piece of hardware is the go-for choice. Will you publish all the Holomark results at some point?

Thank you so much for the effort you put in this matter!

I could easily test the CPU with out using the GPU, and some tests are pure CPU tests, but only doing that would not be a real life test, so I chose to do it the way I did.

And I agree that Nurbs are more important than Meshes, for many, but I think it is important to show how much faster some of the graphiccards are on pure meshes, as this shows the true power of the cards.

I do agree though that the Holomark2 score is not a true indication of how good the system is for average Rhino use, but then again what is an average use? Some users only work with meshes (3D scanning, terrain analysis etc).

And Yes, I will publish the data one of these days!

I guess I have to agree that our use case (nurbs modeling) doesn’t neccessarily reflect the average use case.

On a different note: I reinstalled the FirePro driver for the D300 including restart but the total score is stuck with 23500. Strange, but I think I will leave it at that and return the MacPro.

holomark2test.txt (1.9 KB)

I have got nearly the same configuration…
Nice boost between FirePro M6100 and the “old” FirePro M6000

00 …

Is there a searchable list of results anywhere? One of my GTX 285s is starting to crap out, and I need to replace it (would love to bench mark the other 285 for fun before pulling them, but I see the Holomark link has expired).

Thanks,
Sam

No, not yet. I have to put one together, and will soon.

The current RC2 will expire May 31st so be quick.
I have not figured out why on some systems/tests the red graph is missing. It never happens to me, so this will not be fixed for the release version. I will make it in the beginning of the week, and put together the scores and adjust the “syntethic average” values so the graph don’t jump out of the sheet on the fastest systems :smile:

So stay tuned next week! And have fun in the mean time.

Not bad for such an old bird :wink: (might be time for more than a video card upgrade though)

Holo2Results.txt (2.0 KB)