Network rendering set up Tweet

[cjadams]

This has been like 10 years in the making it seems ever since Nvidia was booted off the mac platform!

So now until apple catches up with nvidia say M4/M5 or M6 era.. then we can ditch the PC too maybe..

But for now.. best of both worlds… working on mac using the power of the RTX 4090 or multiple if your that lucky.. up to 10 render nodes with multiple RTX cards is possible… OH… also best part..this is also fully used when doing IPR functions as well so we have a network rendering accelerated IPR for Lookdev too!…. Combine this with my Stream-deck XL (Look for the shortcut templates over at https://www.sideshowfx.net/products)
They have them pre done for cinema 4d as well as many other apps!

Add to this my 3d spacemouse for navigation.. It is absolutely wonderful being fully on mac with no compromise!

[MShelton]

That is great! Yeah I'm a mix of RTX 4090's and RTX 8000, some RTX 6000's as well in the workstations. Not going to be rendering shots for a few months still so it's mostly a lot of R&D, trying to break things. Mostly render out of Houdini & Maya ( I am a fan of C4D, it's come a long way over the years and is pretty legit).

[cjadams]

Now with scene nodes and such. its on its way to becoming a “Houdini lite” or more friendly accessible platform…. adding in things like x-particles and all its Fused add-ons… and such.. its really nice and well rounded setup!

[Weirdpictures]

Hi Michael,
please consider that distributed rendering solutions can only be used on final rendering, while the beauty of Octane Network rendering is that the Network power can be used in any phase, from scene editing to final rendering, speeding up a lot the entire workflow, and also on single frame rendering with huge resolution.

About license, yes correct, no problem, you can have one primary Workstation + 10x Render-Node machines with a single Studio+ license, and there is no GPU limits per machine!

While with distributed rendering solutions, like Deadline or Tram Render, you need one full Studio+ license per machine involved.

In your case, you could have 2x Studio+ license, one for preparing the scenes (possibly with one or more Render-Nodes), and another one to be use for final rendering with big Network rendering power.

ciao,
Beppe

Hi Beppe, do you know if it is possible to also use just remote nodes for rendering without using the actual workstation that send the job? We have a project where we want to keep working on our main workstation and send the render jobs off to our two connected render stations. They show up perfect in network settings and render perfectly in unison when we start our render job from the workstation. But then our workstation also starts rendering. Is there a way to start teh job from our workstation but have only our network nodes work on the rendering job? thanks in advance for any help.

[MShelton]

@ WeirdPictures-
That's a great question. Would love to hear that answer as well.

Best,
Michael

[bepeg4d]

Hi Beppe, do you know if it is possible to also use just remote nodes for rendering without using the actual workstation that send the job? We have a project where we want to keep working on our main workstation and send the render jobs off to our two connected render stations. They show up perfect in network settings and render perfectly in unison when we start our render job from the workstation. But then our workstation also starts rendering. Is there a way to start teh job from our workstation but have only our network nodes work on the rendering job? thanks in advance for any help.

Hi,
basically you need 2x Studio+ license.
One license is used on the primary machine to prepare the scene, export it in .orbx format, and continue working on other stuff.
The second license is used to run OctaneRender Standalone on a different machine, to become the Workstation at which other Render-Nodes can be connected to, and for rendering the .orbx scenes prepared with license one.

ciao,
Beppe

[cjadams]

that’s kind of weird I swear that I was able to at one point disable the node on the local machine while still using the remote machine for rendering at some point

[Weirdpictures]

Hi Beppe, do you know if it is possible to also use just remote nodes for rendering without using the actual workstation that send the job? We have a project where we want to keep working on our main workstation and send the render jobs off to our two connected render stations. They show up perfect in network settings and render perfectly in unison when we start our render job from the workstation. But then our workstation also starts rendering. Is there a way to start teh job from our workstation but have only our network nodes work on the rendering job? thanks in advance for any help.

Hi,
basically you need 2x Studio+ license.
One license is used on the primary machine to prepare the scene, export it in .orbx format, and continue working on other stuff.
The second license is used to run OctaneRender Standalone on a different machine, to become the Workstation at which other Render-Nodes can be connected to, and for rendering the .orbx scenes prepared with license one.

ciao,
Beppe

Hi Beppe and thanks for your input. But this kind of defeats the whole purpose of having remote rendering network capabilities. I mean if I need a second workstation, then what's the point. I basically can then also end my work on my main workstation, save for example my C4d project and send that via my ethernet connection to one of our render nodes. Open the project there and start the rendering on the gpu farm directly with the second farm acting as a node to that one and not enabling our original workstation. This is exactly the workflow we are doing right now, but it actually always takes a lot of time to do all the above steps, when in reality, the network rendering could be doing this very easy and elegant. All they would need to add is one little menu button in the network render tab that would enable or disable the local machine GPU when starting a network render job. How amazing would that be???

[Weirdpictures]

Oh, just thinking about it, would it maybe be possible to have 2 GPU in our main workstation and when setting up the octane settings to disable the one card for rendering ? So that when we start a network render we have that one GPU disabled? Maybe we could then start another instance of C4D and work with the not rendering card in parallel while the second card is rendering the job together with the network nodes??

Did anyone experiment with that kind of a workflow?

Thanks

[bepeg4d]

Hi,
please consider that also system RAM and CPU are used during rendering, not only the GPU, so better to have a separated Workstation where to send the scenes for final rendering in Net render.

ciao,
Beppe