16

After visiting this page: http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-mac-os-x/

I don't know if my MacBook Pro is suitable for CUDA developement. After About This Mac I see this:

enter image description here

It does not seems to be a graphic card in this list: https://developer.nvidia.com/cuda-gpus

But I asked several times before buying this machine if it is capable for gpu stuffs. And I can correctly runs program that uses shaders. And, more important, I have installed CUDA driver and toolkit and they doesn't seems to give any error. But If I try to compile manually something with CUDA.. it doesn't work.

Improve this question
3
  • 1
    You can run CUDA in software mode, so that the code will be executed by your i5 CPU. This will not be very fast, but it might be enough to learn your first steps with CUDA. But note that the CPU behaves a little bit different from the GPU. So getting another machine with an NVIDIA GPU will be a good idea.
    – Arne
    Commented Jun 7, 2013 at 12:35
  • !! how can i do this?? please send me some link for emulating cuda with the CPU
    – nkint
    Commented Jun 7, 2013 at 12:37
  • Oh, I just noticed that the device emulation mode seems to be deprecated in recent CUDA versions. Back in CUDA 3 this used to work. Sorry for getting your hopes up. The alternative seems to be gpuocelot, but that's Linux only so far. Sorry...
    – Arne
    Commented Jun 8, 2013 at 13:48

5 Answers 5

Reset to default
9

Your model has only the Intel HD 4000, so you can't use CUDA

Full specifications: http://support.apple.com/kb/sp649

Improve this answer
2
  • is there any possibility to change the graphic card or not?
    – nkint
    Commented Jun 7, 2013 at 10:15
  • sorry but is not possible
    – Guido Preite
    Commented Jun 7, 2013 at 10:16
8

About dialog displays only the primary built-in device not the discrete, accelerated GPU. To be sure, run:

system_profiler SPDisplaysDataType

in the terminal.

On my mac, the About dialog shows only on-chip Intel device, but the above command reveals the detail, also telling me that my Mac is equipped with the AMD Radeon HD 6750M which is OpenCL 1.1 capable. My mac is year older then yours.

All MBP devices have two cards.

Graphics/Displays:

Intel HD Graphics 3000:

  Chipset Model: Intel HD Graphics 3000
  Type: GPU
  Bus: Built-In
  VRAM (Total): 512 MB
  Vendor: Intel (0x8086)
  Device ID: 0x0126
  Revision ID: 0x0009
  gMux Version: 1.9.24
  Displays:
    Color LCD:
      Display Type: LCD
      Resolution: 1920 x 1200
      Pixel Depth: 32-Bit Color (ARGB8888)
      Main Display: Yes
      Mirror: Off
      Online: Yes
      Built-In: Yes

AMD Radeon HD 6750M:

  Chipset Model: AMD Radeon HD 6750M
  Type: GPU
  Bus: PCIe
  PCIe Lane Width: x8
  VRAM (Total): 1024 MB
  Vendor: ATI (0x1002)
  Device ID: 0x6741
  Revision ID: 0x0000
  ROM Revision: 113-C0170L-573
  gMux Version: 1.9.24
  EFI Driver Version: 01.00.573
Improve this answer
1
  • very nice answer thanks. and.. yes, only intel is showed
    – nkint
    Commented Nov 11, 2013 at 22:21
1

I recently investigated the possibility of using the Macbook Retina as a CUDA development platform for scientific computing but came away disappointed with it's performance (as of Sep 22, 2014). I had purchased the more expensive model with the Nvidia graphics card for this specific purpose. See Macbook Pro as a CUDA dev (not deployment) platform? for details:

Conclusions

The Nvidia GT 750M card on the 15” Macbook pro Retina running Mac OS X 10.9.5 is not a great CUDA development/proofing platform if the user is interested mainly in double precision, floating point operations. This finding is in direct contrast to claims such as http://blogs.nvidia.com/blog/2012/06/14/new-macbook-pros-make-for-great-cuda-dev-platforms/

Improve this answer
0

Click on the System Report button and navigate to Hardware -> Graphics/Displays. If there's a discrete GPU that matches https://developer.nvidia.com/cuda-gpus, then you've met the System Requirements for CUDA.

Improve this answer
-4

CUDA is only available on nVidia GPUs.

Improve this answer
1
  • 3
    This answer adds nothing that hasn't already been said - in far greater detail.
    – Tetsujin
    Commented Jul 7, 2017 at 17:41

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .