This post should actually be rather short. I will consider what I mean by the term "render farm" as a precursor to the act of displaying data to a screen.



The internet = a network " out there "
An intranet = a network " in here " ( in this case, data transfer would be limited to the computers within one's studio )
Render farm = intranet = a network
SVG is the best way to present graphic data across a network.
Unfortunately, it takes time to render ( mathematically calculate ) and then print the information to a computer monitor. Time needs increase as one zooms an image. Sure, SVG files are smaller than bitmaps. But, it takes a while to render that data to a screen.



How can that be improved?
CPU processing speed is a hardware design issue.
Boards have just so many wires into the CPU.
Even the best task management algorithms cannot improve the speed of a --- render and print to screen sequence --- beyond the potential of a render farm intranet. The problem is the number of wires into and out of the CPU. Adding wires = adding boards = render farm. This is basically all a render farm does for a system.
[ For those who do not understand the CPU and board --- let me use a metaphor - think of a big jug - you want to pour water into it - you have only one size funnel - if you pour too quickly you get the error of a spill - pouring too slow wastes time - you can not use a different funnel because the hole to the jug is just so large - how then do you store more water in the same amount of time? Get another jug and pour into both. CPUs are a big jug for electrons - the funnel and opening are the insertion points where the wires transferring data ( water ) enter the CPU - the only way to store and process more data, when the openings are already full, is to pour into another CPU. One can create a special funnel which has a single " in hole " and several " out holes " which can each be placed into its own jug. Time management won't help. ]



SVG - being for the internet - is already constructed to facilitate network distribution.
Simply develop inkscape to allow distribution of that data over an intranet for calculation and data re-composition before the print to screen.



Say there is a determined threshold of 1 million calculations. At that number, say it takes five minutes to calculate the render and print it to the artist's screen.
The system can be designed so that it knows, at 500,000 calculations - to call upon the resource of a second board. Both boards calculate the render - the data is composited - data is printed to the monitor - the artist saves 2.5 minutes. Being scalable, the render farm ( an intranet ) can be built with a total of 5 boards. Render time just went from 5 minutes to only 1 minute.
( I know the numbers are not actually as clean as I am presenting them, but many readers will probably appreciate the simplicity. )



I actually remember a time when computers had both a motherboard and a baby board.
At first, users will probably need to manually connect their primary computer to a second one through the Ethernet or other port. If the speed becomes popular, hardware manufacturers may return to a miniaturized version of the multi-board design.
As the internet grows - more people will task it to display more graphic data on the internet and such - they will experience slow speeds - this can be a solution.
SVG is being developed, in part, to solve these types of issues.
The work should actually enhance not only the artist's experience but also the potential of the web itself.
Tanks
James
PS - I just know someone is going to tell me that water will make mother boards and computer hardware short circuit.
