SimplyRhino has been testing the real world performance advantage of the new Rhino v6 WIP display pipeline with four up-to-date NVIDIA cards, using a large and highly detailed model of an industrial truck.
If you are interested in reading the report, sign up here to download the full report PDF.
Why only Quadros? It would be great see how they perform compared to GeForce. Let’s give power to the people, not just the elites that can afford Quadros!
As you know, GeForce cards are intended for DirectX gaming applications and nVidia has intentionally throttled them back when they detect graphics calls they interpret as being OpenGL calls.
How instructive would these tests really be?
Just the same, I hope they do test them to get some baseline numbers.
I know what’s Nvidia’s approach and party-line regarding which cards are for which users. I still think all that BS is sometimes irrelevant in real world scenarios. So knowing how well cards that do not cost an arm and a leg do would be great for Rhino customers.
Even for performance and high-end oriented users this information would be very useful. The next system I want to build is a 4X-GPU with GeForce 1080 TIs. Those are th best price/performance for GPU rendering and I do not want to mix-match cards with one Wuadro because that usually messes shit up. It would be good to know how the viewport will behave too, that’s all.
Drop a line to Simply Rhino and see if they have plans to do any testing for V6 on the GeForce cards.
Also. keep in mind that the developers are still working on the display stuff for V6. They are testing on incomplete software.
Totally agree. It is almost impossible to find GTX cards compared to quadro. I understand the arguments about throttling etc but it would still be nice to know the data.
Situation we are in: have two machines with Quadro 4000 cards (old I know). Would like to know if spending £200 on a much newer GTX card will result in a big improvement. Not interested in spending £2000!
yes, for viewport performance the quadros show better performance numbers cause their drivers allow them to utilize all their power.
but when it comes to rendering, all GPU based renderers I know of don’t care if it is a quadro or a gforce - as long as it’s cuda capable.
of course the hardware of quadro is exactly the same as geforce, just another cooler slapped on to it and much more expensive.
look at all the people building octane render workstations - they use lots of GTX1080 or GTX1080Ti and don’t care about quadro.
I would’t be surprised if nvidia made the decision to prohibit this by driver as well in the future.
btw:
nice that cycles devs made the effort to also make it work with openCL
Oh yeah, i’ve found some horrible shading bugs this week. Slits on my model like if it was cut by a fruit ninja. Are you guys aware of these, I should I share it with someone?
It’s ALWAYS best to assume we don’t know about any problem and post a new message with the details, an image, a file, and the steps needed to reproduce it.