I have been teaching Cell processor programming (in C) for the past couple of months. This is definitely not the best place to start, because for successful programming Cell requires mastering many skills: C / C ++, pthreads, libspe, various types of communication in the cell (DMA, mailboxes, signals, interrupts, Atomic I / O). To make this harder, the documentation for Cell can be cryptic, hard to find, and erroneous. If you use a more common platform (XNA, pyGame, SDL), there will be a much larger user community. This does not mean that there is no user community for Cell, but it is less. Although there are other environments where you can find multiprocessor programming, it can be difficult to translate the methods for these environments into a cell due to its unique architecture. In addition, using the standard PS3 with linux will not allow you to access the graphics hardware.
But this is not all bad. Learning the PS3 / Cell will teach you a lot of programming close to the machine. You really have no choice, as there are not many abstractions for the programmer. Each SPU on the cell has 256 Kbytes of local memory, and if you need more than that, you will need to figure out some scheme to issue the correct DMA requests to correctly enter the correct values ββinto the memory at the right time and (hopefully) keep the SPU busy then busy while this DMA request is in flight. Cell study
So, maybe not the best learning platform, but considering that you have the opportunity to take classes from Sony, that sounds good.
In any case, if you are interested, the book from Scarpino is an excellent reference and contains a couple of chapters on game programming on a cell with an OGRE engine, which may also be of interest to you.
Paul wicks
source share