Rhino6 – Quadro P1000 vs GTX 1050ti

unhandled

(bastian) #1

Hey,

i am torn between two Thinkpads (P1/X1E) with pretty identical specs, but coming with these two different GPUs at similar prices (p1 1900€, X1, 2020€). As I am exclusively using Rhino, VRay and AutoCAD with it, I am wondering if – from McNeels side – there is an expectable difference in performance between the two GPUS. Notably, I am confused referring to these two earlier posts, which somehow state contradictory info on OpenGL performance, etc with Rhino6:


What would you say? Any experience with any of these two?

Thanks!
Bastian


#2

Hello,

Probably a lot depends on the tasks you use these programs for.
I use Rhino to create and process nurbs surfaces for reverse
engineering. I work mainly with wireframe and shaded views, without
rendering. I have two computers, one with GTX 650 TI, the other with the
quadro M2000. There is no significant difference between the performance
of these cards in Rhino 6. Probably in AutoCad it will be, but I do not
make major projects on it, so it’s difficult to say. As in previous
Rhino versions most depends on amount of RAM, and single core
performance.

Grzegorz


(bastian) #3

Thanks Gregorz, which that being said I might aswell take the cheaper one? Would be hoping that the 120 extra CUDA cores on the 1050ti wouldnt have made a difference, that cannot be outweighed by optimized drivers and a slightly faster single core performance of the i5-8400. From what Ive read so far, AutoCAD doesnt really care for the GPU anyways, especially as I use it for 2D there


#4

As I wrote earlier - everything can depend on what tasks you use the
graphics card for. After reading many internet forums, I was not sure if
the Quadro gives a real advantage over GTX, and you know what I did? I
bought a cheap, used quadro k600 and tested it in various programs that
I use. Since then I know that cheaper GTX is enough for me, but how will
it be in your case, with other tasks, in Rhino 6, which is a new program, I can not say.

Grzegorz


(bastian) #5

Thanks for the hint, and yes, testing it would actually be the best, but unfortunately not possible in my case (will be a new workstation, and reseller doesnt offer returns for businesses). After all, I assume that it will hardly make any noticable difference at all, so testing will be judging on very subjective grounds.

To be honest, I like the idea of Quadros priotizing stability over performance, as with occassional renders it wont hurt to much if it takes longer, but I experienced system crashes delaying my workflow to great extents.

Still, would have been interesting to hear sb. from McNeel on the compatibiliy of both cards with Rhino6 (and to what extent the new render and viewport favours OpenGL over DirectX cards), as that was something that startled the confusion for me, at least – any opinions on that McNeel?


(John Brock) #6

We do not test display adapters to the extent you would need for your comparison.


(Wim Dekeyser) #7

Also see this reply for some general advice:


(Kyle Houchens) #8

raytraced mode does use cuda cores- more= better


(bastian) #9

To all, thanks for your feedbacks. Actually, both are pretty much based on the same hardware, and have 4GB of VRAM (8 are out of budget unfortunately), yet the Quadro seems downclocked for stability and offers only 512 vs 768 cuda cores. I mostly use wireframe, but naturally, better performance is a pro.

I just thought there is any rational and proven argument why McNeel recommends Quadros – as system req. calls them the best and pro choice for Rhino. If not, one has nothing but Cudas to ground a decision upon, and that again makes the 1050ti a clear winner


(Rocchi33) #10

watch this video of the time 15 minutes ahead and would like the opinion of you.
there it shows that the pictures are not good for the cudas, but for the way they work differently.