I’m looking at the new macbook pro’s from apple… and thinking of doing some c for cuda/opencl programming on it.
I’ve been looking at the newest programming guide for cuda and the chips 9600m GT and 9400m are not listed for computability levels… is that becourse they are not supported? Have anyone tried running cuda code on them… or is there a place to look for specs on the chips… and not least specs on the drivers…
thank you
Martin
As some of the latest Nvidia chipsets they should support CUDA which came in with the G80 series…
I’ll let you know definitively very soon as I have a mac book on order for this very purpose…
It definitely supports CUDA with 9400M, because I’m using macbook 466 right now, with 9400M nvidia card working fine with CUDA
thats very interesting, what sorta performance are you getting on that machine?
And what are the memory limits, is it possible to store large tings in the “graphics mem”? And even more interesting is copying from the host memory to the device mem faster than usual?
I’m extremely interested :)
Martin
New MacBook with 9400M works! Initially it didn’t work, but then I came across another thread that mentioned that the Toolkit installer doesn’t install an essential kext unless it’s selected via the “customize” button. Now everything is peachy.
I’ve got new MacBook (white) + 9400M, 2GB DDR2, OSX 1.5.6 + CUDA Toolkit 2.0.
– A
Wikipedia has a relatively detailed list of specs. I’m maxing out at 25 GFLOPS/sec on my MacBook running the nbody sample app with n=65536.
– A