in the last times I heard lots of people claiming that the Cell processor is dead, mainly due to the following reasons:
a) lack of support in the new playstation 3, as the user can not install linux
b) the increasing processing power of the GPU's and its costs sinking
c) the existence of a unified programming approach (openCL) for different GPU's and not for the CBE (well today was announced for the Cell!)
d) carency of real world examples of use of the cell (apart from the academic circles)
e) global feeling of unsuccess
What do you think? If you started two or three years ago to program the cell, will you continue on this or are you considering switching to GPU's? Is a new version of the cell coming?
a) The removal of Linux support on the PS3 Slim is a shame, but anyone who was serious about coding on Cell started off on the simulator or QS22 anyway.
b) GPUs and the Cell are not directly comparable - GPUs are still very domain-specific due to the architecture, and if they were any more generic, they'd be CPUs and have the same limitations (or they'll end up with something that looks very much like Cell!).
c) Yes, OpenCL was announced for Cell.. so you can see development is still going on!
d) Real world examples, whilst there aren't that many outside of gaming and academia there are some:
e) Has it achieved the kind of numbers everyone hoped, outside of gaming? Unfortunately not, but still you'd be hard pressed to find a comparable processor that has both the power and the general applicability of Cell, regardless of it being difficult to code for it is still easier than the alternatives. The only contenders (and feel free to correct me) are Tilera (no FPU on the cores, only 50 GFLOPS and no real world samples yet), and GPUs (still very specific problem domains and even more difficult to code for but, if the shoe fits, great performance) at this stage. Are there others?
There are things happening behind closed doors on Cell, so I wouldn't say it was dead. It's just never been marketed correctly.
