Thinking about getting a dedicated computer for doing renderings in your office? There are a couple of considerations that you need to be aware of before doing so.
Estimating user load
When estimating required performance, you need to consider the number of users that will utilize the service.
Please keep in mind that this service was originally designed to be installed on most computers in an office to share idle computing power, not just one or two. Although you can certainly do this, there are a couple of factors that you need to keep in mind.
Avoiding queue build-up
To avoid the risk of a queue build-up, you need to consider how often your individual users will render an image. This in combination with how many users will send render jobs to a specific machine determines performance load. It might even be better to set up multiple, less powerful, machines to help alleviate the risk of queue build-up.
This is because the service works on a first come, first serve basis, meaning that the service will render one job at a time. As the service does not currently load balance or split performance between multiple jobs, this means that a queue of requested jobs can build up when multiple users wish to have their images rendered at the same time.
All rendering is done on the computer’s CPU, so there is no need for an extravagantly powerful graphics card.
The more and the faster CPU cores, the quicker a render job will be completed. You will want to ensure at lease 8 GB of RAM, though more is better.
The original design idea called for multiple computers on the same network to all have the render server installed, thus sharing idle computing power. When taking this into account, it may be better to dedicate two or more medium level computers to this task, rather than one high performance computer.