Forum: 3D Modeling


Subject: Some modeling

LuxXeon opened this issue on Mar 09, 2017 ยท 29 posts


LuxXeon posted Sat, 22 April 2017 at 4:00 PM

HMorton posted at 3:18PM Sat, 22 April 2017 - #4303043

I hear all professional users always talking about Quadro cards. Even when I look at Boxxx workstations, they come with Quadro. But why is it that when it comes to gpu rendering, I always hear everyone saying to go with a high end gtx card instead? Even in this thread I noticed that oldguy has a quadro in his machine, but luxxeon said he's using a gtx. So what's the difference, and why are Quadro's almost double the cost of most gtx cards? I'm asking because I want to get a new system for gpu rendering, and want to choose the right Nvidia card to get the fastest results. If anyone has any insight or can share their own story about why they use the nvidia card they use, that would be great!

Sorry for the late reply, HMorton. Quadro cards are typically found in CAD workstations like the Dell Precision series. and are specifically engineered to use specific drivers that work best with certain specifications in professional CAD software. Quadro cards are made to handle large polygon datasets in demanding software, and perform this task much better than a GTX card would. Quadro cards usually support higher onboard memory (VRAM), and support larger screen resolutions with support for 30-bit color. The extra memory allows you to navigate very large 3D scenes with ease and supports multiple displays at much higher resolution. Obviously, if you are doing architectural development or visualization this would be a very important thing.

The difference between a Quadro M6000 ($5000) and Nvidia's powerful Titan X (around $1200) card is the memory support. Quadro M6000 has between 12- 24gigs of DDR5 GPU ram. As you can imagine, this is very handy when working on scenes with possibly millions of polygons in a CAD application.

On the other hand, Quadro cards do not have the rendering power of high-end gaming cards like the GTX. The Titan X, for example, has 3,583 CUDA cores on board. CUDA cores are used by all GPU render engines to speed up the renders, so the more cores your card has, the faster it will render a scene. By comparison, the Quadro M6000 has 3,072. So the Titan X has 300+ more cores for rendering than the M6000. This means that when you send your scene to a render engine like Octane or Cycles, the Titan X would render the scene faster than the Quadro. This is comparing two of the highest end Nvidia cards, but the overall comparison holds true for other cards in the two lines as well.

If your primary concern is viewport performance and memory handling while working on large, complex polygon scenes, then Quadro cards would be the best way to go. On the other hand, if you're buying a card specifically for GPU-accelerated rendering with CUDA, then a GTX or Titan card would have more CUDA cores and thus provide faster renders.

______________________________________

My Store
My Free Models
My Video Tutorials
My CG Animations
Instagram: @luxxeon3d
Facebook: https://www.facebook.com/luxxeon