Breaking News: On stage at the GTC2010 Keynote, NVidia announced that they are working with PGI, the Portland Group, to develop a new compiler CUDA-x86. If you don’t have a GPU cluster handy, you can run your CUDA code directly on the CPU for testing and debugging, and even deploy it to users who don’t have suitable GPU’s.
Of course, you’ll need the GPU for optimum performance, but it makes CUDA not just an NVidia technology, but a technology that runs on any computer.
Update 12:10pm it was confirmed later that this will not be a free offering, but rather a commercial product marketed and sold by PGI.
Update 9/23 1pm: More information here.
@ someone Doh. I’m wrong. It seems it’s strictly a CPU-only compiler. Read the details here.
A researcher already posted a simple script that converts your cuda code to fast multithreaded cpu code. I use it all the time. Also you could just use OpenCL so I don’t understand what the big deal is.
The main difference is that the decision to execute on the GPU or Cpu is made at runtime, not build
Time (as I understand it). It’s an important distinction as it means a single complex application can run on any computer, but offer a nice boost if you have a CUDA capable card.