c++ - Cuda compilation of examples

Question

I have NVIDIA GeForce 8500 GT on Windows 7 Pro 32, and I have problem with my project in CUDAC. I have installed all packages and VS2012 Pro. I'm creating new Project from template for Cuda 6.5... Compile it and.. "invalid device function". From Geting Started Windows PDF I have read that i can chceck CUDA by deviceQuery.exe.. So i done this :

deviceQuery.exe Starting...

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

Detected 1 CUDA Capable device(s)

Device 0: "GeForce 8500 GT"
  CUDA Driver Version / Runtime Version          6.5 / 6.5
  CUDA Capability Major/Minor version number:    1.1
  Total amount of global memory:                 512 MBytes (536870912 bytes)
  ( 2) Multiprocessors, (  8) CUDA Cores/MP:     16 CUDA Cores
  GPU Clock rate:                                1570 MHz (1.57 GHz)
  Memory Clock rate:                             400 Mhz
  Memory Bus Width:                              128-bit
  Maximum Texture Dimension Size (x,y,z)         1D=(8192), 2D=(65536, 32768), 3D=(2048, 2048, 2048)
  Maximum Layered 1D Texture Size, (num) layers  1D=(8192), 512 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(8192, 8192), 512 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       16384 bytes
  Total number of registers available per block: 8192
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  768
  Maximum number of threads per block:           512
  Max dimension size of a thread block (x,y,z): (512, 512, 64)
  Max dimension size of a grid size    (x,y,z): (65535, 65535, 1)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             256 bytes
  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  CUDA Device Driver Mode (TCC or WDDM):         WDDM (Windows Display Driver Model)
  Device supports Unified Addressing (UVA):      No
  Device PCI Bus ID / PCI location ID:           1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 6.5, CUDA Runtime Version = 6.5, NumDevs = 1, Device0 = GeForce 8500 GT
Result = PASS

So PASS!!! SO WHATS WRONG..? Next i done bandwidthTest

[CUDA Bandwidth Test] - Starting...
Running on...

 Device 0: GeForce 8500 GT
 Quick Mode

 Host to Device Bandwidth, 1 Device(s)
 PINNED Memory Transfers
   Transfer Size (Bytes)    Bandwidth(MB/s)
   33554432         1346.5

 Device to Host Bandwidth, 1 Device(s)
 PINNED Memory Transfers
   Transfer Size (Bytes)    Bandwidth(MB/s)
   33554432         1556.9

 Device to Device Bandwidth, 1 Device(s)
 PINNED Memory Transfers
   Transfer Size (Bytes)    Bandwidth(MB/s)
   33554432         5857.4

Result = PASS

So can enybode help my?

Answer 1

Invalid device function usually means that the code was compiled with an architecture that is higher than the GPU you are trying to run it on.

The GPU architecture is contained in your printout:

  CUDA Capability Major/Minor version number:    1.1

CUDA 6.5 compiles for a cc2.0 architecture by default. If you want to compile for a cc 1.1 architecture, you will need to pass specific switches to your nvcc compile command to do so.

This usually means adding something like compute_11,sm_11 in the Visual Studio device configuration tab on your project properties.

When you do so, you will then get warnings (under CUDA 6.5) that device architecture 1.1 is deprecated. However you can still compile for and target this architecture.

And even though this question pertains to windows, the same necessity exists on Linux. If you use CUDA 6.5 on linux, the default compile target is cc2.0. To compile for an earlier device, it's necessary to add something to the compile command line like -arch=sm_11.