Holomark 2 for Rhino 5

Thanks, and you are welcome!

Many users are willing to spend 10x the price of Rhino on the graphiccard alone (Quadro K6000) so knowing what, or even if, one gain on such an investment I believe is important.

And thanks for the advice @donl517, I’ll look into Google form. If you find some info please share.

Now off to the garden and the tree cottage building with the kids :smile:

@MattE @Holo Here is your screencast: Holomark Screencast

If it spun my models as fast as Jorgen’s, I would love this card!

What do you mean by that? You got better scores than I did.

Your car spins around really fast with no hesitation, but when i try and spin my models around it is real laggy.


Head Slap, So thinking about my earlier comment I wondered about Jorgen’s display modes vs. the ones I have been using. I have had the “Backface Settings” Set to: “Single Color for all backfaces” I set that back to the defaults and now I can spin my part without lag.

@jeff Why would this cause so much lag? Aren’t the backfaces being created/generated regardless in either case?


Hi Don, that doesn’t make sense to me… I must missunderstand something, because you get better scores than I do, and if you compare my youtube video to your screencast then you will see that your video is less laggy. More JAGGY, but less laggy.

So excuse my lack of understanding and please explain in detail exactly what you experience as less laggy.


Well just try this, take a large model, set backfaces to a single color for all backfaces and than try and manipulate the model view. Then set it back to use front face settings and try it again.

See if this helps explain what I was seeing: Lag in rotation

A new GPU test perhaps?

I’ll gladly try that, but what am I looking for, or comparing to? What do you see?

Ah, you just edited the post with a link. I’ll look!

Is it a low frame rate, or does it “hang” for a while?
Is it a lagg between changing display modes you are experiencing?

If you give me a test, by using the mini, and your preferred display mode, then I can time it. Even make a script for you to compare to my result if you like.


EDIT: I see the lagg. BUT I still don’t get your comparison, where did you experience that my system was smoother or less laggy than yours?

EDIT 2: MY HEAD SLAP this time! I didn’t see that you referred to my vs your MODEL… Giiz. Yeah, and this is why Holomark2 installs custom display modes, so it doesn’t compare your edited modes to my modes :smile:

Now that we are on the same page, it might be worth a look at how this effects different systems. Honestly I never gave it a thought, until today.


1 Like

I continue to be thoroughly excited by the prospect of an updated Holomark. Thanks for doing all this work Holo! (and thanks to beta testers and others pitching in).

Thinking about all the time (and money) I’ve spent over the last… ten years(!) trying to figure out what’s the best video performance (for the least $)… anyway, Holomark 2 is going to be awesome!

1 Like

I agree, thanks to all the beta testers!
I hope to release RC1 next week, with some minor improvements and bug fixes.

And BTW the current Beta stops working on May 1st.

Hi Holo, I’m moving on up!

I built a new system, with a similar processor to Don’s, just a newer chip (4820). I held off on the purchase of a V7900 and settled for a avarage Graphics Card, until the dust settles on which Graphics Card is the “New Champion”.

I went from a grand total of 23,000 to 30,576.

See attached:
HoloMark2-03.txt (2.0 KB)


Yes, back faces are created and being drawn regardless… However, a series of polygons can only be drawn a certain way at any given time… which means if you have a series of polygons that all look the same, then they can all be drawn at the same time. If some polygons are supposed to look different than others, then it will require a multi-step process (Draw some polygons one way, then draw other polygons a different way).

Since you had backfaces being drawn different than front faces, that meant all front faces need to be drawn using one material, and then all backfaces drawn using a different material…thus, really drawing each object twice (but not really quite that bad).

I just did a test on the mini file in rendered mode.
TestMaxSpeed took 6.18 seconds.

Then I toggled on the custom color for backfaces and
TestMaxSpeed took 6.41 seconds.

So that is far from a performance hit.

BUT I messed around a bit more and turned on isocurves and surface edges:
With OUT backface color: 10.11 sec
WITH backface color: 8.4 sec

Still not a huge performance difference, but this time it goes the other way around.

So if you export your display mode setting @donl517 then I can test the speed difference on my system.
And please try this on your model too.

I created a separate display mode for back faces, but these settings are pretty vanilla.

I should note the file I’m working with is 204MB.

Don_Shaded.zip (4.6 KB)

Ok, and what are you TestMaxSpeed scores on your file with those two display modes? Is it a big difference?
TestMaxSpeed is a command you type and then Rhino rotates the view 5 degrees 100 times and returns how long it took.
It is a wee bit faster than the command I use in Holomark2.

I’ll check later.

9.14 in shaded, 27.41 in backface.


@donl517 Can you run GPU-Z at the same time to see what your video memory usage is like?

Yeah, that was a huge performance drop on your system.
I just tested on a 53MB hybrid file with mainly meshes on my laptop:

Command: Testmaxspeed
Time to regen viewport 100 times = 12.56 seconds.
Display mode set to “Backfaces”.
Command: Testmaxspeed
Time to regen viewport 100 times = 12.90 seconds.

So not much here, and this is with a GeForce 330M

Video Usage memory never moves off of 415MB in either display mode, running TestMaxSpeed, GPU load is at 45% in BackFace, and flickers to 16% for a fraction of a second in Shaded (The rest of the time it is 0%).