Monday, July 29, 2013

Propeller Supercomputer

Evolution of the super propeller stage
Some thoughts on building a Propeller supercomputer

Don't give up on your projects. Why? They said it couldn't be accomplished. They bickered and argued, flamed and incensed their tempers and raised blood pressure. Not standard they yelled. Impossible they cited. Never been done before they reminded. They demanded stoppage. It's too power demanding. Too costly. Too big. Too crazy. No useful purpose. No apps. They insisted it was pure senseless madness.

The questions asked included, would it be fast enough, would it have enough processing power, would it have enough memory, would it do this, would it do that, would it float, would it...? With too many arguable questions, too many debates, too many mouths speaking all at the same time, too many repeats, and not enough action, we decided to move forward.

To gain enough power, we concatenated processors. To gain enough speed, we added GPUs exceeding Teraflops. To gain enough resources, we added memory, terabyte drives, monitors, cards, and additional computers. To gain enough computational thinking power, we paralleled the circuits. To gain software flexibility, we accessed the ultimate list of propeller languages and ran a fractional number of the near 260 possibilities.

It didn't end there. But to make a long story short, after years of development, the Propeller Supercomputer is complete and accomplished. So if you have a project and you hear the infamous words, "it can't be done," just move forward and finish it. Never give up on it and don't listen to the nay-sayers.