Sounds good, but from what I gained from the Kickstarter, you only get 2 ARM A9 chips and 16 of their cores.

Even if it has 2 ARM processors, the big deal of this is that maybe we should be able to use the Epiphany coprocessor in a way similar to what GPU workunits do, so when this comes out, and BOINC supports it in the future, imagine having Epiphany workunits available to crunch on that coprocessor. If I compare it to a GPU, I think both would be handled the same way in BOINC, so this should be something to look forward, especially since they claim each core runs at 1Ghz

Hi Gromobir,
Words like Balkanization come to mind. So does Ruritania or Lilliputia. I don't mean to sound cryptic. Compiler and library development for new architectures or new bundles made up of older architectures is not easy or trivial. Then using it in a distributed fashion requires publicity to somehow reach the few. We have very few old projects that can still run on the Power-PC architecture. Everything else is x86. We are breaking into Windows x86 with OpenCL (GPU) but still have Linux x86 and Mac OSX x86 to go.

The great new architectures of the past, like the Transputer, look like traps for software developers today.

Lawrence

@lawrencehardin: Ok, I get your point and I know that it is a hell of a work to port so many software projects to a new architecture. But on the other hand, it's something necessary in my opinion. If there is a superior new technology we should definitefily embrace it and not stick with the older one because of compatibility reasons.
But I guess it would be with this as it was with other older architectures: The hobbyists would start working on the platform first before any respectable company would get on the radar. But someone has to start.