fresh install is worse 
then my oldest laptop…
any help on this?
Hi Jennifer,
Are you actually experiencing slowdowns in real operation (not Holomark)?
For example just spinning a heavy model around on the screen (your impression of speed/fluidity as compared to the other computers you have)
Or doing some complex boolean operation - does it perform more slowly than the others?
–Mitch
yes yes everything is slow on this new machine
grasshopper can take ages just to calculate a volume
booleens even simple ones take longer than the old computers
I am lost…
Merde, alors…
Hi Jennifer,
you could reinstall the Nvidia driver, best use the Nivida driver uninstall tool before:
Select the option “new installation” (or called like this, I have a german UI only)
Would a video driver affect boolean operation speed? --Mitch
Micha, thanks that was my last resort I will try that maybe next week because I am going away for work… but all the tests I did on the gtx were fine… I will do that but the cpu seems slow as well but only in rhino…
after new driver on CG Nvidia Thanks Micha
But still under performing
Any other suggestions?
If I compare the result with my GTX780 results I would say the CPU could the problem. For example my wireframe test is faster, but my shaded test is slower.
You could update the chipset driver. Maybe both -GPU and CPU driver - needs to be at the same “time” level.
I updated my chip set driver and now the calculating of the technical display is 10 times slower. I don’t see why it is so slow now. Other CPU and GPU values are like before, no change, so why the calculating of the technical display?
Before:
Holomark 2 v2,2,03
Total Score: 31666
Total Runtime: 139.53 sec
GPU scores: 19560
GPU_01 - 851.10 fps - Cube 4 tests
GPU_02 - 34.10 fps - UDT Shape
GPU_03 - 81.30 fps - Wireframe
GPU_04 - 40.30 fps - Shaded
GPU_05 - 29.70 fps - Rendered
GPU_06 - 29.20 fps - Block Rendered
GPU_07 - 17 units Nurbs @ 5 fps in Wireframe
GPU_08 - 9 units Nurbs @ 5 fps in Shaded
GPU_09 - 10 units Nurbs @ 5 fps in RenderSpeed
GPU_10 - 92.60 fps - RenderMesh Render
GPU_11 - 69.00 fps - RenderMesh RenderSpeed
GPU_12 - 128.20 fps - JoinedMesh Render
GPU_13 - 54.60 fps - JoinedMesh RenderSpeed
GPU_14 - 2 units mesh @ 15 fps in Shaded
GPU_15 - 6 units mesh @ 15 fps in Render
GPU_16 - 4 units mesh @ 15 fps in RenderSpeed
GPU_17 - 80.00 fps - mesh in Rendered Studio
GPU_18 - 9.30 fps - Nurbs in Rendered Studio
GPU_19 - 20.50 fps - Block Illustration
GPU_20 - 47.20 fps - 2D single
GPU_21 - 5.70 fps - 2D massive (20x)
CPU scores: 12106
CPU_01 - 10.83 sec - Booleans and Contours
CPU_02 - 1.67 sec - Twist and Taper (UDT)
CPU_03 - 4.74 sec - Meshing Mini
CPU_04 - 0.02 sec - Extract Render Mesh
CPU_05 - 0.09 sec - Join Render Mesh
CPU_06 - 17.44 sec - Reduce Mesh
CPU_07 - 3.03 sec - Calculating Technical display
CPU_08 - 3.48 sec - Making Silhouettes
ASUSTeK COMPUTER INC.
Z9PE-D8 WS
NVIDIA GeForce GTX 780 - 4095.0 MB
DriverVersion: 9.18.13.4725
After:
Holomark 2 v2,2,03
Total Score: 27420
Total Runtime: 184.8 sec
GPU scores: 21480
GPU_01 - 1025.60 fps - Cube 4 tests
GPU_02 - 34.20 fps - UDT Shape
GPU_03 - 80.00 fps - Wireframe
GPU_04 - 41.20 fps - Shaded
GPU_05 - 30.50 fps - Rendered
GPU_06 - 29.70 fps - Block Rendered
GPU_07 - 17 units Nurbs @ 5 fps in Wireframe
GPU_08 - 9 units Nurbs @ 5 fps in Shaded
GPU_09 - 10 units Nurbs @ 5 fps in RenderSpeed
GPU_10 - 92.60 fps - RenderMesh Render
GPU_11 - 69.40 fps - RenderMesh RenderSpeed
GPU_12 - 133.30 fps - JoinedMesh Render
GPU_13 - 55.20 fps - JoinedMesh RenderSpeed
GPU_14 - 2 units mesh @ 15 fps in Shaded
GPU_15 - 6 units mesh @ 15 fps in Render
GPU_16 - 4 units mesh @ 15 fps in RenderSpeed
GPU_17 - 85.50 fps - mesh in Rendered Studio
GPU_18 - 9.00 fps - Nurbs in Rendered Studio
GPU_19 - 20.70 fps - Block Illustration
GPU_20 - 51.30 fps - 2D single
GPU_21 - 5.90 fps - 2D massive (20x)
CPU scores: 5940
CPU_01 - 10.59 sec - Booleans and Contours
CPU_02 - 1.83 sec - Twist and Taper (UDT)
CPU_03 - 4.65 sec - Meshing Mini
CPU_04 - 0.02 sec - Extract Render Mesh
CPU_05 - 0.08 sec - Join Render Mesh
CPU_06 - 17.11 sec - Reduce Mesh
CPU_07 - 46.38 sec - Calculating Technical display
CPU_08 - 3.51 sec - Making Silhouettes
ASUSTeK COMPUTER INC.
Z9PE-D8 WS
NVIDIA GeForce GTX 780 - 4095.0 MB
DriverVersion: 9.18.13.4725
Hmm, I don’t understand much here… My latest score with my home machine is 71395… I tested previously with the same setup and it was 45844… (post 252). What’s going on there? Only difference, SR11 RC (latest) instead of SR10.
Is it possible there is that much difference or is the test unreliable? --Mitch
OK, that wasn’t the latest SR11, it was the previous one. Loaded the latest and now it’s back down to 42869… So what changed? Again, how reliable is the test really?
Yeah, something is really wrong here:
GPU 13 1666.7 --> 54.30
GPU 16 180 units --> 4 units
Don’t believe that loading the new RC could change that much.
This thread here is a bit related :
Holomark’s reliablilty should be good as the scores are repeatable, but it may not reflect real life scenarios as it can be a script -> plugin issue, but this is just guessing wildly, as I have nothing to back up the “idea”.
That the machine runs that fast with Maxwell rules out any irregularities from say the memory or the machine as that.
Have you tried to run Rhino bare of all Plug-ins? You could also check in the task manager, what else is going on while Rhino is running. I presume, you are not at the internet with the work machine. Else or anyway I would first do a thorrough virus check.
It simply can’t be that your exemplar of Rhino should behave other than others on an otherwise normal machine.
You could also download another example of the installation file, maybe it happened that once in a billion chance, that there crept in a bug via download, that lets Rhino run, but not good enough. I know, it sounds like bullshit, but…
steff
Are you sure, that your AA settings were identical?
Ah yeah. And the 6gb for gpu rendering are a good reason too.
Maybe it is the C in RC. You don’t know what the girls and guys changed. Why should it be Holo, when all it does is aline valid commands?
But on a sidenote, and I believe that’s the language in what it is written, maybe a bit timing sensitive. I noticed rather often a stop in the test right before gpu16. I mean, there might be a command still be processed, though the cursor is already back at the command line for the next command, and therefore it sometimes stumbles. Just my laymans idea about it.
And, yes, I also noticed jumps of quite a percentage in the results of certain disciplines of the test. But nothing in the magnitude of your above mentioned results.
Below I uploaded two runs under identical settings. GPU_13 eg jumps at about 30%, which is quite much I imagine. Must be something to do with the number of samples considered for the run. I can’t express that correctly, I am not a math guy.
GPU_14 jumps even more, with more than two thirds of a hundred percent.
Well, CPU_05, that is not that clear, with the resultion showed. First result could be up to 0.045, and the second down to 0.0251.
Not a great score, but for a tiny portable computer I won’t complain. It’s actually totally usable for simple modeling and script development.
Correction: Hardware acceleration was off the first time, GPU results are much better now.
Not bad for integrated graphics, I’d say.
Hmmm… Apparently not, no.
Of course, http://www.rhino3d.com/download/rhino/5.0/sr/rc/changes, should tell us exactly that but checking this one against the RC1 log would lead us to believe that nothing has been changed between those two.
Steff
Yes for all but the recent virus check
I will try that
Micha: where do you find the cpu update Intel is a maze
Tx