PDA

View Full Version : Small renderfarm: Satellite or Standalone? Faster?


benvu
11-29-2004, 04:23 AM
I am planning to use four of my dual CPU workstations as a xsi mr renderfarm (more like a rendergarden :)). One or two artist working at a time, so small operation going. Which would be more ideal and faster for distributed rendering? Satellite or Standalone?

Vertizor
11-29-2004, 05:13 AM
Not entirely sure what you mean by "stand alone" but with satellite, that helps with rendering single frames. 4 dual CPU machines for a total of 8 CPUs, and 2 artists working equates to 4 CPUs per artist. When working on static images, each artist gets 2 computers to themselves. But for animations, but all 4 of them to work.

juanjgon
11-29-2004, 08:11 AM
I have a question about this topic, so i cant test it ... the four satellite nodes available in essential license work with dual CPUs render node to complete a total of 8 aditional CPUs, or are this nodes only available for one CPU node?

benvu
11-29-2004, 08:27 AM
We are planning out the renderfarm now. Have not completed purchase yet. Yes, we would want to use every CPU available so we would need two-node licenses. Need to know which licensing method (Satellite or Standalone) would work faster for animation rendering.

According to XSI docs:
How Distributed Rendering Works Distributed rendering is started automatically once a render is initiated on a computer. The initiating computer is referred to as the master and the other computers on the network are referred to as slaves. The master and slaves communicate via a mental ray service that listens on a designated TCP port and passes information to mental ray.

Distributed Rendering Types: Satellite vs. Standalone

You can configure two types of distributed rendering: Satellite and Standalone.



Satellite distributed rendering requires an XSI license that supports Satellite rendering for the master machine. No additional licences are required for the slave machines. Instead, the master has a fixed number of Satellite tokens, each of which can be used for a render slave processor. On multi-processor machines, each processor requires a separate token. The number of available Satellite tokens varies depending on which type of XSI license you have.
Satellite distributed rendering only works when rendering using XSI (interactively or from the command line).


Standalone distributed rendering requires an XSI license that supports distributed rendering for the master machine. Each slave processor requires an additional rendering license (XSI|Batch or XSI|Batch Universal). On multi-processor machines, each processor requires a separate license.
Standalone distributed rendering works when rendering using either XSI or the mental ray standalone.


During the setup process, you can configure one or both types of distributed rendering. Each type uses its own service that listens on its own port. Keep in mind, however, that they cannot be used simultaneously. If you’ve configured both Satellite and Standalone distributed rendering for a single setup, you’ll need to choose one of them prior to rendering. You can do this by editing the .ray3hosts file, which lists the render slaves.

Vertizor
11-29-2004, 03:22 PM
Just to make it short and sweet: use satellite for rendering still frames, stand alone for animations.

In all cases you'll need Mental Ray licenses for each CPU. One seat of Essentials just cover the workstation the artist works on. You'll need 4 more MR CPU licenses to cover the 2 other slave machines. The Satellite licenses simply allow you to talk to the slaves from the master, it's not an MR license.

The real difference between satellite vs standalone is: satellite breaks up the frame into regions and each region is a "job" distributed to a node. Standalone distributes whole frames (of an animation) to each node as a "job."

CGTalk Moderation
01-20-2006, 12:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.