Announcing OctaneRender Cloud Edition

[RayTracey]

Hi everyone,

OTOY is proud to announce its GPU cloud rendering service for Octane Render called Octane Render Cloud Edition, which was officially presented today during the opening keynote of Nvidia's GTC conference.

Below is a video fragment of the GTC opening keynote showing Octane Render on the cloud in action (the Octane Render demo was rendering remotely on 112 GPUs simultaneously)



Higher quality video: http://www.ustream.tv/recorded/30095793

The pricing of the service is announced at a maximum of $1/GPU hour, with 4-128 GPUs per user on demand. The cost may be lower during non-peak hours.
The service will initially be available in the US only during the beta period (starting in July) with international support planned for later this year.

Cheers,

The OctaneRender team

[pixelrush]

Are those gpu Teslas? How much vram then? 4/6 gb? And how is the scaling with that many gpu?
How long queue time do you anticipate or are these to run in off hours? How many gpu is one job allowed? Stills only or animations too?
What is the least practical render time or number of samples with 128 gpu then considering the tasks of loading and rendering? All those gpu together would be able to spit out finished renders every few seconds wouldnt they? Do we pay for the load time or just the render time?
Ha lots of questions...
Good development. I am impressed the way Otoy are moving ahead with Octane.

[PAQUITO]

This is amazing. I would like a service to send my scene to, setted up to a number of samples, then receiving back the final image and paying pre-paid credits. That´s the way Rebusfarm works, and it´s incredibly practical.

[juanjgon]

Amazing !!!

-Juanjo

[face_off]

Fantastic news. Congratulations to Otoy.

Paul

[RayTracey]

Are those gpu Teslas? How much vram then? 4/6 gb? And how is the scaling with that many gpu?
How long queue time do you anticipate or are these to run in off hours? How many gpu is one job allowed? Stills only or animations too?
What is the least practical render time or number of samples with 128 gpu then concidering the tasks of loading and rendering? All those gpu together would be able to spit out finished renders every few seconds wouldnt they? Do we pay for the load time or just the render time?
Ha lots of questions...
Good development. I am impressed the way Otoy are moving ahead with Octane.

The GPUs are custom Kepler based cloud GPUs from Nvidia with 4 GB each. The scaling is perfect, i.e. rendering a still or animation on 128 GPUs simultaneously will be exactly 128x faster than on one GPU (we did many tests to confirm this). Queue time should be nil, details on the cloud service will be revealed soon. You can render on 4 to 128 GPUs on demand and you can use them for whatever you want to render, stills or animations. We have put a lot of time and effort into optimizing the scene loading and updating across the network to get immediate feedback from all the GPUs. Loading a scene on all the GPUs of the rack happens almost instantly as all nodes are interconnected with ultra high speed connections. A few seconds of rendering time are sufficient to get noise free renders at very high resolutions (it's very satisfying to see the progress bar fill up to 16000 samples/px using PMC in just a few seconds ). We will only charge for actual rendertime.

[pixelrush]

Wow! Great!
OK so I presume this 'Cloud Edition' is for the farms only.
So how will users make use of it? Will there be something incorporated in a future Octane UI or will there be a small app for that? Sorry if I am asking too many questions, perhaps I should just wait for the announcement

[bepeg4d]

impressed
a new starting point in the 3d visualization... sooner or later we will have octane on tablet
ciao beppe

[MOSFET]

Very, very cool Ray. Congratulations to the Octane team

[kavorka]

I've been waiting for this news since summer!

amazing!

Will this work with any version of Octane, standalone and plugin?