Nvidia Reveals $50,000 Iray VCA Rendering Accelerator
Nvidia introduced a new appliance for rendering.
At Nvidia's GTC 2014 conference, the company introduced a GPU rendering appliance called the Iray VCA (Visual Computing Appliance). This appliance aims to eliminate the costly, lengthy process of building physical prototypes by rendering computer models with extremely high visual fidelity.
Nvidia's Iray is a photo-realistic renderer already used in Autodesk's 3ds Max and a few other tools. Not only does the appliance accelerate the renderer, but it's scalable, meaning a number of appliances can be linked together, speeding up the overall rendering by "hundreds of times."
"Iray VCA lets designers do what they've always wanted to -- interact with their ideas as if they were already real," said Jeff Brown, vice president and general manager of Professional Visualization and Design at Nvidia. "It removes the time-consuming step of building prototypes or rendering out movies, enabling designs to be explored, tweaked and confirmed in real time. Months, even years -- and enormous cost -- can be saved in bringing products to market."
According to the specs given during the keynote, this appliance costs $50,000. The device includes eight Kepler-based GPUs, 12 GB of memory per GPU, and 23,040 CUDA cores. The ingredients also consist of two 1 GigE ports, two 10 GigE ports, and one InfiniBand connection.
Nvidia reports that Iray VCAs can be located in an IT center and serve up rendering power on demand "while requiring little or no technical support." Honda is already one of the early adopters, with a prototype cluster made up of 25 nodes to refine styling designs on future cars.
The Iray VCA will be made available this summer through CADnetwork, Fluidyna, IGI and migenius. The cost includes an Iray license and the first year of maintenance and updates.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
There's a budget GeForce GPU selling in China that not even Nvidia knew it made — RTX 4010 turns out to be a modified RTX A400 workstation GPU
US to patch loopholes that allow China to buy banned AI GPUs from other countries — new regulations include national quotas on GPU exports and a global licensing system
-
Damon Palovaara Wow I might have to sell my house for this, who needs a big house anyways when you have one of these. Hell you can live in your own virtual mansion if you wanted toReply -
Free2play_noobs @Damon Palovaara Obviously it is meant for enterprise use ,not for personal/Consumer use.Reply -
rwinches Far cry from the really cool state of the art Silicon Graphics SGI workstations of old.__Honda bought 25 nodes so this is good for Nvidia, I certainly want competition for AMD ATI.Reply -
The device includes eight Kepler-based GPUs .... and 23,040 CUDA cores
So that is 2880 CCs/GPU, meaning 8 Tesla K40 running in parallel? Wow, that's more than 11 TFLOPS of double precision power in that box. -
eodeo Thats equivalent of 8x 780TIs with 12gb of ram each. That's actually pretty good value, if on the expensive side for small studios."I certainly want competition for AMD ATI."Unfortunately, AMD cannot fight back as nvidia bought mental images, creators of mental ray and used it as base for iray. Iray is essentially mental ray ported to work on CUDA. There is no AMD alternative planed.Reply -
mr meeseeks That is not an image of the Iray VCA. He is holding the compute module for Audi's self driving cars that fit's in a slot in the trunk. The VCA is sized to be rackmounted.Reply -
outlw6669 This appliance aims to eliminate the costly, lengthy process of building physical prototypes by rendering computer models with extremely high visual fidelity.
Sure.... unless you actually need to test the prototype.
You know.... like you need to do with just about every prototype....
While a higher fidelity CAD model might be nicer for some; in the end a model is just a model is just a model.
If you are building anything functional, you will still need to have at least one physical prototype phase to prove your system.