CUDA driver on Mac OS Mavericks

Hi guys, I have a macbook with a GeForce 320M graphic card. I was programming using CUDA SDK with Mountain Lion and latest CUDA driver without any problems, but since I installed Mavericks OS the CUDA SDK and driver seems not to be working any more. I get the following error when I try to execute a cuda example:

cudaGetDeviceCount returned 38
→ no CUDA-capable device is detected

I searched on the web for a little bit of information and there is some people with the same problem after upgrade to Mavericks. I’m not quite sure if the problem is with the CUDA driver, the Mavericks OS or both. Hope anyone could help with this or give me more information about this subject. Thanks for your help!

Cheers

Exactly the same problem on Macbook Air 2010. I’m sure that bgonzales has either white Macbook or Macbook air 2010 model.

I posted this problem last week, but my post was deleted, I hope this one will not be deleted. Same problem here, GT330M, CUDA 5.5.28 and OS X Mavericks

I get the same issue on a late 2008 Macbook with 9400M graphic card.

In my case, I found it might be because my card is a 1.1 CUDA capability card and NVIDIA stated in latest release notes (http://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html) :

Nevertheless, I didn’t check if your 320M or 330M are 1.x CUDA capability cards.

Hoooo, yes indeed our cards are something like 1.2 or 1.3 capable, so it explains the problems. I hope they will release the drivers soon.

Thanks

Yes, please release drivers soon. The Adobe forums are filled with people upgrading to Mavericks, just to find CUDA crippled. Lucky for me I attempted to upgrade on a fresh SSD and had my prior system intact.

This may be unrelated, I believe 10.9 also supports OpenGL 4.xx, will this have any positive impact?

Thanks!

Same Problem here with my 9400M/9600M GT… waiting for a release. (had a lot of searching till I found the answer…)

Same issue here with cuda-mac-5.5.28_10.9_64.pkg and NVIDIA GeForce GT 330M 512 MB. :-(

Same issue here:
OS X 10.9
CUDA Driver Version: 5.5.28
GPU Driver Version: 8.18.27 310.40.05f01

Will this be resolved any time soon?

Hi, I’ve just installed CUDA 5.5.28. Same problem.
OS X 10.9
CUDA Driver Version: 5.5.28
GPU Driver Version: 8.18.27 310.40.05f01

./deviceQuery Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount returned 38
→ no CUDA-capable device is detected
Result = FAIL

any workaround until the new driver release ?

Same problem here, late 2008 macbook with 9400M.
OS X 10.9
CUDA Driver Version: 5.5.28
GPU Driver Version: 8.18.27 310.40.05f01

./deviceQuery Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount returned 38
→ no CUDA-capable device is detected
Result = FAIL

… so are we still waiting for a driver to be released? Just upgraded a 2010 Macbook Pro 13" to OSX 10.9, and CUDA is no longer recognized. Trying to decide if I should just go back to Mountain Lion?

Any updates on this?

Same problem here. Please restore CUDA functionality for my GeForce GT 330M under OS X Mavericks.

yes,indeed our cards are something like 1.2 or 1.3 capable

/Developer/NVIDIA/CUDA-5.5/samples/bin/x86_64/darwin/release/deviceQuery Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount returned 38
→ no CUDA-capable device is detected
Result = FAIL

::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

[oclDeviceQuery] starting…

/Developer/NVIDIA/CUDA-4.2/samples/OpenCL/bin/darwin/release/oclDeviceQuery Starting…

OpenCL SW Info:

WARNING: NVIDIA OpenCL platform not found - defaulting to first platform!

CL_PLATFORM_NAME: Apple
CL_PLATFORM_VERSION: OpenCL 1.2 (Aug 24 2013 21:03:27)
OpenCL SDK Revision: 7027912

OpenCL Device Info:

2 devices found supporting OpenCL:


Device Intel(R) Xeon(R) CPU E5520 @ 2.27GHz

CL_DEVICE_NAME: Intel(R) Xeon(R) CPU E5520 @ 2.27GHz
CL_DEVICE_VENDOR: Intel
CL_DRIVER_VERSION: 1.1
CL_DEVICE_VERSION: OpenCL 1.2
CL_DEVICE_TYPE: CL_DEVICE_TYPE_CPU
CL_DEVICE_MAX_COMPUTE_UNITS: 8
CL_DEVICE_MAX_WORK_ITEM_DIMENSIONS: 3
CL_DEVICE_MAX_WORK_ITEM_SIZES: 1024 / 1 / 1
CL_DEVICE_MAX_WORK_GROUP_SIZE: 1024
CL_DEVICE_MAX_CLOCK_FREQUENCY: 2260 MHz
CL_DEVICE_ADDRESS_BITS: 64
CL_DEVICE_MAX_MEM_ALLOC_SIZE: 3072 MByte
CL_DEVICE_GLOBAL_MEM_SIZE: 12288 MByte
CL_DEVICE_ERROR_CORRECTION_SUPPORT: no
CL_DEVICE_LOCAL_MEM_TYPE: global
CL_DEVICE_LOCAL_MEM_SIZE: 32 KByte
CL_DEVICE_MAX_CONSTANT_BUFFER_SIZE: 64 KByte
CL_DEVICE_QUEUE_PROPERTIES: CL_QUEUE_PROFILING_ENABLE
CL_DEVICE_IMAGE_SUPPORT: 1
CL_DEVICE_MAX_READ_IMAGE_ARGS: 128
CL_DEVICE_MAX_WRITE_IMAGE_ARGS: 8
CL_DEVICE_SINGLE_FP_CONFIG: denorms INF-quietNaNs round-to-nearest round-to-zero round-to-inf fma

CL_DEVICE_IMAGE 2D_MAX_WIDTH 8192
2D_MAX_HEIGHT 8192
3D_MAX_WIDTH 2048
3D_MAX_HEIGHT 2048
3D_MAX_DEPTH 2048

CL_DEVICE_EXTENSIONS: cl_APPLE_SetMemObjectDestructor
cl_APPLE_ContextLoggingFunctions
cl_APPLE_clut
cl_APPLE_query_kernel_names
cl_APPLE_gl_sharing
cl_khr_gl_event
cl_khr_fp64
cl_khr_global_int32_base_atomics
cl_khr_global_int32_extended_atomics
cl_khr_local_int32_base_atomics
cl_khr_local_int32_extended_atomics
cl_khr_byte_addressable_store
cl_khr_int64_base_atomics
cl_khr_int64_extended_atomics
cl_khr_3d_image_writes
cl_khr_image2d_from_buffer
cl_APPLE_fp64_basic_ops
cl_APPLE_fixed_alpha_channel_orders
cl_APPLE_biased_fixed_point_image_formats

CL_DEVICE_PREFERRED_VECTOR_WIDTH_ CHAR 16, SHORT 8, INT 4, LONG 2, FLOAT 4, DOUBLE 2


Device GeForce GT 120

CL_DEVICE_NAME: GeForce GT 120
CL_DEVICE_VENDOR: NVIDIA
CL_DRIVER_VERSION: 8.18.28 310.40.05f01
CL_DEVICE_VERSION: OpenCL 1.0
CL_DEVICE_TYPE: CL_DEVICE_TYPE_GPU
CL_DEVICE_MAX_COMPUTE_UNITS: 4
CL_DEVICE_MAX_WORK_ITEM_DIMENSIONS: 3
CL_DEVICE_MAX_WORK_ITEM_SIZES: 512 / 512 / 64
CL_DEVICE_MAX_WORK_GROUP_SIZE: 512
CL_DEVICE_MAX_CLOCK_FREQUENCY: 1250 MHz
CL_DEVICE_ADDRESS_BITS: 32
CL_DEVICE_MAX_MEM_ALLOC_SIZE: 128 MByte
CL_DEVICE_GLOBAL_MEM_SIZE: 256 MByte
CL_DEVICE_ERROR_CORRECTION_SUPPORT: no
CL_DEVICE_LOCAL_MEM_TYPE: local
CL_DEVICE_LOCAL_MEM_SIZE: 16 KByte
CL_DEVICE_MAX_CONSTANT_BUFFER_SIZE: 64 KByte
CL_DEVICE_QUEUE_PROPERTIES: CL_QUEUE_PROFILING_ENABLE
CL_DEVICE_IMAGE_SUPPORT: 1
CL_DEVICE_MAX_READ_IMAGE_ARGS: 128
CL_DEVICE_MAX_WRITE_IMAGE_ARGS: 8
CL_DEVICE_SINGLE_FP_CONFIG: INF-quietNaNs round-to-nearest round-to-zero round-to-inf

CL_DEVICE_IMAGE 2D_MAX_WIDTH 4096
2D_MAX_HEIGHT 4096
3D_MAX_WIDTH 2048
3D_MAX_HEIGHT 2048
3D_MAX_DEPTH 2048

CL_DEVICE_EXTENSIONS: cl_APPLE_SetMemObjectDestructor
cl_APPLE_ContextLoggingFunctions
cl_APPLE_clut
cl_APPLE_query_kernel_names
cl_APPLE_gl_sharing
cl_khr_gl_event
cl_khr_byte_addressable_store
cl_khr_global_int32_base_atomics
cl_khr_global_int32_extended_atomics

CL_DEVICE_PREFERRED_VECTOR_WIDTH_ CHAR 1, SHORT 1, INT 1, LONG 1, FLOAT 1, DOUBLE 0


2D Image Formats Supported (28)

Channel Order Channel Type

1 CL_INTENSITY CL_FLOAT
2 CL_RGBA CL_FLOAT
3 Unknown CL_FLOAT
4 Unknown CL_FLOAT
5 Unknown CL_FLOAT
6 CL_INTENSITY CL_HALF_FLOAT
7 CL_RGBA CL_HALF_FLOAT
8 Unknown CL_HALF_FLOAT
9 CL_RGBA Unknown
10 CL_RGBA CL_SIGNED_INT16
11 CL_RGBA CL_SIGNED_INT32
12 CL_RGBA CL_SIGNED_INT8
13 CL_RGBA CL_SNORM_INT16
14 CL_RGBA CL_SNORM_INT8
15 CL_RGBA CL_UNSIGNED_INT16
16 CL_RGBA CL_UNSIGNED_INT32
17 CL_RGBA CL_UNSIGNED_INT8
18 CL_RGBA CL_UNORM_INT16
19 Unknown CL_UNORM_INT16
20 Unknown CL_UNORM_INT8
21 CL_A CL_UNORM_INT8
22 CL_ARGB CL_UNORM_INT8
23 Unknown CL_UNORM_INT8
24 CL_BGRA CL_UNORM_INT8
25 CL_INTENSITY CL_UNORM_INT8
26 CL_RGBA CL_UNORM_INT8
27 Unknown CL_UNORM_INT8
28 Unknown CL_UNORM_INT8


3D Image Formats Supported (28)

Channel Order Channel Type

1 CL_INTENSITY CL_FLOAT
2 CL_RGBA CL_FLOAT
3 Unknown CL_FLOAT
4 Unknown CL_FLOAT
5 Unknown CL_FLOAT
6 CL_INTENSITY CL_HALF_FLOAT
7 CL_RGBA CL_HALF_FLOAT
8 Unknown CL_HALF_FLOAT
9 CL_RGBA Unknown
10 CL_RGBA CL_SIGNED_INT16
11 CL_RGBA CL_SIGNED_INT32
12 CL_RGBA CL_SIGNED_INT8
13 CL_RGBA CL_SNORM_INT16
14 CL_RGBA CL_SNORM_INT8
15 CL_RGBA CL_UNSIGNED_INT16
16 CL_RGBA CL_UNSIGNED_INT32
17 CL_RGBA CL_UNSIGNED_INT8
18 CL_RGBA CL_UNORM_INT16
19 Unknown CL_UNORM_INT16
20 Unknown CL_UNORM_INT8
21 CL_A CL_UNORM_INT8
22 CL_ARGB CL_UNORM_INT8
23 Unknown CL_UNORM_INT8
24 CL_BGRA CL_UNORM_INT8
25 CL_INTENSITY CL_UNORM_INT8
26 CL_RGBA CL_UNORM_INT8
27 Unknown CL_UNORM_INT8
28 Unknown CL_UNORM_INT8

oclDeviceQuery, Platform Name = Apple, Platform Version = OpenCL 1.2 (Aug 24 2013 21:03:27), SDK Revision = 7027912, NumDevs = 2, Device = Intel(R) Xeon(R) CPU E5520 @ 2.27GHz, Device = GeForce GT 120

System Info:

[oclDeviceQuery] test results…
PASSED

exiting in 3 seconds: 3…2…1…done!

i have a Late 2008 with Mavericks… and the same problem No Cuda Capable Device is found, please fix it. I need Cuda for my university thanks

Pls support your brend in our Macs . Pls restore power of rendering via GPU . Pls start harmony of CUDA and GeForce GT 330M and Mavericks . thx

Hi,

I had the same problem as in the previous posts, but I accidentally find something which might be the solution.

After downloading CUDA 5.5 the driver version on my system was 5.5.28. This is reported by ‘CUDA Preferences’ in ‘System Preferences’. The same dialog reports this driver version as the most up-to-date.

So I downloaded a really old driver (I picked 4.2.10 randomly from [url]https://developer.nvidia.com/cuda-toolkit-42-archive[/url], download driver only) and installed it on my system.

The I went back to the ‘CUDA Preferences’ in ‘System Preferences’ dialog, which reported the 4.2.10 driver version, and offered installing CUDA update. After installing, the version became 5.5.43!

And now deviceQuery works!

Let me know if you find this helpful.
Cheers
Albert

yes, new driver release ! it works!