deviceQuery on 8600GT 512MB + CUDA 하드웨어 구조

deviceQuery 예문은 말그대로 장치에 질의의 던져 어떤 스펙인지
몇개의 thread가 존재하는지를 파악하는 프로그램이다.

아무튼 NVIDIA_CUDA_C_ProgrammingGuide.pdf 문서를 보면 아래의 내용이 나오니 참고를 해서 보자면,

Geforce 8800GT(이하 8800GT)는 14개의 멀티 프로세서를 가졌고 112개의 코어를 지녔다.
Grid는 Block을 포함하고, Block은 Thread를 포함한다.

단순 계산으로는 하나의 프로세서별로 8개의 코어가 존재하며
멀티프로세서는 14개 코어는 총 112개가 존재한다.

그리드(Grid)는 블럭의 2차원 배열로 존재하고,
블럭(Block)은 쓰레드의 2차원 배열로 존재한다.
쓰레드(Thread)는 일을하는 최소단위이다.

한번에 묶이는 최소 쓰레드의 숫자(Warp size)는 32개 이며
하나의 블럭으로 묶을수 있는 최대 쓰레드는 512개 이다.

블럭의 최대 차원은 3차원 512x512x64 이며
그리드의 최대 차원은 2차원 65535x65535x1 이다.

라고 이해하면 되려나?

A multithreaded program is partitioned into blocks of threads that execute independently from each other, so that a GPU with more cores will automatically execute the program in less time than a GPU with fewer cores. => 멀티쓰레드화된 프로그램은 서로 독립적으로 실행되는 쓰레드의 블럭으로 나누어지며, 그러한 이유로 더욱 많은 코어를 포함하는 CPU는 적은 코어를 지닌 GPU보다 짧은 시간에 프로그램을 실행할 수 있다.

Host는 일반적인 CPU 환경이고, Device는 GPU 환경이다.
컴파일시에 Device 코드만 nvcc에서 처리하고, 나머지는 일반적인 컴파일러에서 처리하는 이원화된 구조이다.

아래는 커널이다. 디바이스 코드를 생성하는 부분이며
MatAdd 함수는 __global__ 선언을 앞에 붙여 device 코드임을 명시해야 한다.

MatAdd<<<numBlocks, threadsPerBlock>>>(A, B, C);

그리고 커널 안에는 int형이나 dim3 형으로 선언이 되어야 한다.

아래는 dim3 형으로 선언된 녀석으로,

dim3 threadsPerBlock(16, 16); dim3 numBlocks(N / threadsPerBlock.x, N / threadsPerBlock.y); MatAdd<<<numBlocks, threadsPerBlock>>>(A, B, C);

블럭당 쓰레디의 크기를 16x16 thread로, 그리드를 N/16개로 분할하는 예제이다.

물론 2차원은 1차원과 같이 사용이 가능하므로 int형도 허용하는 듯?

B.15  Execution Configuration Any call to a __global__ function must specify the execution configuration for that call. The execution configuration defines the dimension of the grid and blocks that will be used to execute the function on the device, as well as the associated stream (see Section 3.3.10.1 for a description of streams). When using the driver API, the execution configuration is specified through a series of driver function calls as detailed in Section 3.3.3. When using the runtime API (Section 3.2), the execution configuration is specified by inserting an expression of the form <<< Dg, Db, Ns, S >>> between the function name and the parenthesized argument list, where:   Dg is of type dim3 (see Section B.3.2) and specifies the dimension and size of the grid,      such that Dg.x * Dg.y equals the number of blocks being launched; Dg.z must be equal to 1;   Db is of type dim3 (see Section B.3.2) and specifies the dimension and size of each block,      such that Db.x * Db.y * Db.z equals the number of threads per block;   Ns is of type size_t and specifies the number of bytes in shared memory that is dynamically allocated per block for this call in addition to the statically allocated memory; this dynamically allocated memory is used by any of the variables declared as an external array as mentioned in Section B.2.3; Ns is an optional argument which defaults to 0;   S is of type cudaStream_t and specifies the associated stream; S is an optional argument which defaults to 0. As an example, a function declared as __global__ void Func(float* parameter); must be called like this: Func<<< Dg, Db, Ns >>>(parameter); The arguments to the execution configuration are evaluated before the actual function arguments and like the function arguments, are currently passed via shared memory to the device. The function call will fail if Dg or Db are greater than the maximum sizes allowed for the device as specified in Appendix G, or if Ns is greater than the maximum amount of shared memory available on the device, minus the amount of shared memory required for static allocation, functions arguments (for devices of compute capability 1.x), and execution configuration.

아무튼 원래는 아래의 예제 내용을 분석하기 위한 내용인데
점점 미궁으로 빠지는 느낌 -_-

CUDA deviceQuery

D:\CUDA\NVIDIA GPU Computing SDK\C\bin\win32\Release\deviceQuery.exe Starting...  CUDA Device Query (Runtime API) version (CUDART static linking) There is 1 device supporting CUDA Device 0: "GeForce 8800 GT"   CUDA Driver Version:                           3.20   CUDA Runtime Version:                          3.10   CUDA Capability Major revision number:         1   CUDA Capability Minor revision number:         1   Total amount of global memory:                 536543232 bytes   Number of multiprocessors:                     14   Number of cores:                               112   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 block:           512   Maximum sizes of each dimension of a block:    512 x 512 x 64   Maximum sizes of each dimension of a grid:     65535 x 65535 x 1   Maximum memory pitch:                          2147483647 bytes   Texture alignment:                             256 bytes   Clock rate:                                    1.50 GHz   Concurrent copy and execution:                 Yes   Run time limit on kernels:                     Yes   Integrated:                                    No   Support host page-locked memory mapping:       Yes   Compute mode:                                  Default (multiple host threads can use this device simultaneously)   Concurrent kernel execution:                   No   Device has ECC support enabled:                No deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 3.20, CUDA Runtime Vers ion = 3.10, NumDevs = 1, Device = GeForce 8800 GT PASSED Press <Enter> to Quit... -----------------------------------------------------------

OpenCL DeviceQuery

D:\CUDA\NVIDIA GPU Computing SDK\OpenCL\bin\Win32\Release>oclDeviceQuery.exe oclDeviceQuery.exe Starting... OpenCL SW Info:  CL_PLATFORM_NAME:      NVIDIA CUDA  CL_PLATFORM_VERSION:   OpenCL 1.0 CUDA 3.2.1  OpenCL SDK Revision:   6161726 OpenCL Device Info:  1 devices found supporting OpenCL:  ---------------------------------  Device GeForce 8800 GT  ---------------------------------   CL_DEVICE_NAME:                       GeForce 8800 GT   CL_DEVICE_VENDOR:                     NVIDIA Corporation   CL_DRIVER_VERSION:                    260.99   CL_DEVICE_TYPE:                       CL_DEVICE_TYPE_GPU   CL_DEVICE_MAX_COMPUTE_UNITS:          14   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:        1500 MHz   CL_DEVICE_ADDRESS_BITS:               32   CL_DEVICE_MAX_MEM_ALLOC_SIZE:         128 MByte   CL_DEVICE_GLOBAL_MEM_SIZE:            511 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_OUT_OF_ORDER_EXEC_MODE_ENABLE   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 fma   CL_DEVICE_IMAGE <dim>                 2D_MAX_WIDTH     4096                                         2D_MAX_HEIGHT    32768                                         3D_MAX_WIDTH     2048                                         3D_MAX_HEIGHT    2048                                         3D_MAX_DEPTH     2048   CL_DEVICE_EXTENSIONS:                 cl_khr_byte_addressable_store                                         cl_khr_icd                                         cl_khr_gl_sharing                                         cl_nv_d3d9_sharing                                         cl_nv_compiler_options                                         cl_nv_device_attribute_query                                         cl_nv_pragma_unroll                                         cl_khr_global_int32_base_atomics                                         cl_khr_global_int32_extended_atomics   CL_DEVICE_COMPUTE_CAPABILITY_NV:      1.1   NUMBER OF MULTIPROCESSORS:            14   NUMBER OF CUDA CORES:                 112   CL_DEVICE_REGISTERS_PER_BLOCK_NV:     8192   CL_DEVICE_WARP_SIZE_NV:               32   CL_DEVICE_GPU_OVERLAP_NV:             CL_TRUE   CL_DEVICE_KERNEL_EXEC_TIMEOUT_NV:     CL_TRUE   CL_DEVICE_INTEGRATED_MEMORY_NV:       CL_FALSE   CL_DEVICE_PREFERRED_VECTOR_WIDTH_<t>  CHAR 1, SHORT 1, INT 1, LONG 1, FLOAT 1,  DOUBLE 0   ---------------------------------   2D Image Formats Supported (71)   ---------------------------------   #     Channel Order   Channel Type   1     CL_R            CL_FLOAT   2     CL_R            CL_HALF_FLOAT   3     CL_R            CL_UNORM_INT8   4     CL_R            CL_UNORM_INT16   5     CL_R            CL_SNORM_INT16   6     CL_R            CL_SIGNED_INT8   7     CL_R            CL_SIGNED_INT16   8     CL_R            CL_SIGNED_INT32   9     CL_R            CL_UNSIGNED_INT8   10    CL_R            CL_UNSIGNED_INT16   11    CL_R            CL_UNSIGNED_INT32   12    CL_A            CL_FLOAT   13    CL_A            CL_HALF_FLOAT   14    CL_A            CL_UNORM_INT8   15    CL_A            CL_UNORM_INT16   16    CL_A            CL_SNORM_INT16   17    CL_A            CL_SIGNED_INT8   18    CL_A            CL_SIGNED_INT16   19    CL_A            CL_SIGNED_INT32   20    CL_A            CL_UNSIGNED_INT8   21    CL_A            CL_UNSIGNED_INT16   22    CL_A            CL_UNSIGNED_INT32   23    CL_RG           CL_FLOAT   24    CL_RG           CL_HALF_FLOAT   25    CL_RG           CL_UNORM_INT8   26    CL_RG           CL_UNORM_INT16   27    CL_RG           CL_SNORM_INT16   28    CL_RG           CL_SIGNED_INT8   29    CL_RG           CL_SIGNED_INT16   30    CL_RG           CL_SIGNED_INT32   31    CL_RG           CL_UNSIGNED_INT8   32    CL_RG           CL_UNSIGNED_INT16   33    CL_RG           CL_UNSIGNED_INT32   34    CL_RA           CL_FLOAT   35    CL_RA           CL_HALF_FLOAT   36    CL_RA           CL_UNORM_INT8   37    CL_RA           CL_UNORM_INT16   38    CL_RA           CL_SNORM_INT16   39    CL_RA           CL_SIGNED_INT8   40    CL_RA           CL_SIGNED_INT16   41    CL_RA           CL_SIGNED_INT32   42    CL_RA           CL_UNSIGNED_INT8   43    CL_RA           CL_UNSIGNED_INT16   44    CL_RA           CL_UNSIGNED_INT32   45    CL_RGBA         CL_FLOAT   46    CL_RGBA         CL_HALF_FLOAT   47    CL_RGBA         CL_UNORM_INT8   48    CL_RGBA         CL_UNORM_INT16   49    CL_RGBA         CL_SNORM_INT16   50    CL_RGBA         CL_SIGNED_INT8   51    CL_RGBA         CL_SIGNED_INT16   52    CL_RGBA         CL_SIGNED_INT32   53    CL_RGBA         CL_UNSIGNED_INT8   54    CL_RGBA         CL_UNSIGNED_INT16   55    CL_RGBA         CL_UNSIGNED_INT32   56    CL_BGRA         CL_UNORM_INT8   57    CL_BGRA         CL_SIGNED_INT8   58    CL_BGRA         CL_UNSIGNED_INT8   59    CL_ARGB         CL_UNORM_INT8   60    CL_ARGB         CL_SIGNED_INT8   61    CL_ARGB         CL_UNSIGNED_INT8   62    CL_INTENSITY    CL_FLOAT   63    CL_INTENSITY    CL_HALF_FLOAT   64    CL_INTENSITY    CL_UNORM_INT8   65    CL_INTENSITY    CL_UNORM_INT16   66    CL_INTENSITY    CL_SNORM_INT16   67    CL_LUMINANCE    CL_FLOAT   68    CL_LUMINANCE    CL_HALF_FLOAT   69    CL_LUMINANCE    CL_UNORM_INT8   70    CL_LUMINANCE    CL_UNORM_INT16   71    CL_LUMINANCE    CL_SNORM_INT16   ---------------------------------   3D Image Formats Supported (71)   ---------------------------------   #     Channel Order   Channel Type   1     CL_R            CL_FLOAT   2     CL_R            CL_HALF_FLOAT   3     CL_R            CL_UNORM_INT8   4     CL_R            CL_UNORM_INT16   5     CL_R            CL_SNORM_INT16   6     CL_R            CL_SIGNED_INT8   7     CL_R            CL_SIGNED_INT16   8     CL_R            CL_SIGNED_INT32   9     CL_R            CL_UNSIGNED_INT8   10    CL_R            CL_UNSIGNED_INT16   11    CL_R            CL_UNSIGNED_INT32   12    CL_A            CL_FLOAT   13    CL_A            CL_HALF_FLOAT   14    CL_A            CL_UNORM_INT8   15    CL_A            CL_UNORM_INT16   16    CL_A            CL_SNORM_INT16   17    CL_A            CL_SIGNED_INT8   18    CL_A            CL_SIGNED_INT16   19    CL_A            CL_SIGNED_INT32   20    CL_A            CL_UNSIGNED_INT8   21    CL_A            CL_UNSIGNED_INT16   22    CL_A            CL_UNSIGNED_INT32   23    CL_RG           CL_FLOAT   24    CL_RG           CL_HALF_FLOAT   25    CL_RG           CL_UNORM_INT8   26    CL_RG           CL_UNORM_INT16   27    CL_RG           CL_SNORM_INT16   28    CL_RG           CL_SIGNED_INT8   29    CL_RG           CL_SIGNED_INT16   30    CL_RG           CL_SIGNED_INT32   31    CL_RG           CL_UNSIGNED_INT8   32    CL_RG           CL_UNSIGNED_INT16   33    CL_RG           CL_UNSIGNED_INT32   34    CL_RA           CL_FLOAT   35    CL_RA           CL_HALF_FLOAT   36    CL_RA           CL_UNORM_INT8   37    CL_RA           CL_UNORM_INT16   38    CL_RA           CL_SNORM_INT16   39    CL_RA           CL_SIGNED_INT8   40    CL_RA           CL_SIGNED_INT16   41    CL_RA           CL_SIGNED_INT32   42    CL_RA           CL_UNSIGNED_INT8   43    CL_RA           CL_UNSIGNED_INT16   44    CL_RA           CL_UNSIGNED_INT32   45    CL_RGBA         CL_FLOAT   46    CL_RGBA         CL_HALF_FLOAT   47    CL_RGBA         CL_UNORM_INT8   48    CL_RGBA         CL_UNORM_INT16   49    CL_RGBA         CL_SNORM_INT16   50    CL_RGBA         CL_SIGNED_INT8   51    CL_RGBA         CL_SIGNED_INT16   52    CL_RGBA         CL_SIGNED_INT32   53    CL_RGBA         CL_UNSIGNED_INT8   54    CL_RGBA         CL_UNSIGNED_INT16   55    CL_RGBA         CL_UNSIGNED_INT32   56    CL_BGRA         CL_UNORM_INT8   57    CL_BGRA         CL_SIGNED_INT8   58    CL_BGRA         CL_UNSIGNED_INT8   59    CL_ARGB         CL_UNORM_INT8   60    CL_ARGB         CL_SIGNED_INT8   61    CL_ARGB         CL_UNSIGNED_INT8   62    CL_INTENSITY    CL_FLOAT   63    CL_INTENSITY    CL_HALF_FLOAT   64    CL_INTENSITY    CL_UNORM_INT8   65    CL_INTENSITY    CL_UNORM_INT16   66    CL_INTENSITY    CL_SNORM_INT16   67    CL_LUMINANCE    CL_FLOAT   68    CL_LUMINANCE    CL_HALF_FLOAT   69    CL_LUMINANCE    CL_UNORM_INT8   70    CL_LUMINANCE    CL_UNORM_INT16   71    CL_LUMINANCE    CL_SNORM_INT16 oclDeviceQuery, Platform Name = NVIDIA CUDA, Platform Version = OpenCL 1.0 CUDA 3.2.1, SDK Revision = 6161726, NumDevs = 1, Device = GeForce 8800 GT System Info:  Local Time/Date = 20:52:54, 1/2/2011  CPU Arch: 0  CPU Level: 15  # of CPU processors: 2  Windows Build: 2600  Windows Ver: 5.1 PASSED Press <Enter> to Quit... -----------------------------------------------------------