I had an idea for a business. Start up with as many nodes as you can purchase, preferable something along the lines of 20-100 basic computers, necessary networking hardware, whatever software is needed, to start a sort of a render farm business, where you'd rent computer time to, say, animation and graphics students who needed to render 3d graphics but don't have access to a good render farm, places that need number processing or distributed computing on a small scale, etc. As you make money, add more nodes as you can afford them, upgrading as necessary (then again, render farms don't really need the best hardware, do they?), advertising...

Is this at all a feasible idea? Likely to make any money? It would certainly require an investment to begin with, probably a small business loan to get the startup hardware and usable premises, but would it be worth it?

Bump.

Y'all are reading this, I can tell. Post, dammit!

probably not, anyone that really needed distributed processing power could just do it on their existing network.

it might work but having better computers will speed up the work and will help "sell" you to people who need that type work

i saw a server farm on Alias ......
are you going to have a detonator for the guards to detonate the sensitive data if there is a breach?




(on a serious note: it is possible, and not difficult to set one up, but why?  if it were a few cray computers it might be worth it....we had server farms in college and universities would share cpu time with other universities for major projects, etc)

Does it have to be a real farm with chickens and Cows?

I suppose you could put the computers in the barn!

How much are people willing to pay for the service?  What sort of software will you need?  What sort of competition will you face?

The computers can be pretty minimal -- no monitor, keyboard, mouse, disk, or optical drives, just boot them off the network -- but you're going to have to find a market for it.  There aren't a lot of people who need a cluster.

You might even need to use some of those "blade" computers in order to keep down costs.  Strangely, the ones I've seen advertised were really high priced, not sure why since they cut out so much of the legacy hardware crap.

How many people really need a beowulf'ed cluster of computers?  Sure, it might be fun.  But I really cant see how it would be profitable.


Now to get your name out, what you want to do is just beowulf something strange, like a cluster of Furbys.

Will they be free range computers because I believe they have the same rights as us.

Let's see. Can you give us a decent estimate on the number of people who might actually use your service?  How many people actually do rendering enough that they would want to rent a system, anyway? And, if they do, why don't they have their own gear since, presumably, they are in the business.

If you don't know that, then you aren't even at the point where you could ask this question.

My first impression is that students don't make good, reliable markets for anything other than music, booze, and drugs. In general, you don't want to pick a major market where nobody in the market really has much money. They would be more likely to use whatever the school provided -- which means you would have to market it to the schools.

First things first --- how many people (total) might have any interest at all in this service?

Second - assuming you knew the first number, how would you propose to reach them with advertising that would  be remotely cost-effective?