NVN wrote:There was a precedent in the year 2012?
The Oracle case I linked to was 2012
bepeg4d wrote:...consider the irrefutable fact that your GPUs are not working while rendering in Cinema 4D CPU.
For now,
GPU rendering is coming native in C4D.
Maybe the GPUs are used for simulation, encoding or another GPU renderer? They're getting plentiful and more ones are on the way, so please consider the irrefutable fact that Octane isn’t the only software that uses the GPU.
bepeg4d wrote:
With a bit different approach, you would be able to render with Cinema 4D and OctaneRender in the same time drastically cutting your rendering time, instead of insist to calculate one by one.
Let’s say I have 15 shots with a total of 30000 frames and render time is 1 min/frame on a 4 GPU machine.
On 30 machines that’ll be almost 17 hours of rendering.
If I instead do it the way you suggest I can use 5 machines before hitting the arbitrary 20 GPU limit. Sure I cut per frame render time to 12 seconds, but total render time will be 100 hours instead of 17 hours. Oh, there’s also a 20 sec preparation time per frame, which brings my 30 machines up to about 22 hours and your approach comes in at something like 260 hours.
In short:
My way - Under a day
Your way - 1 week and almost 3 days
But then again, to render on 30 machines with Octane you’ll need to spend $18000.
That’s more than 20 times as expensive as some of the competition, any ideas how I’m supposed to justify that to my CFO?