Cuda 学习教程二：GeForce GTX 950M 显卡参数

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


Detected 1 CUDA Capable device(s)


Device 0: "GeForce GTX 950M"
  CUDA Driver Version / Runtime Version         8.0 / 8.0
  CUDA Capability Major/Minor version number:      5.0
  Total amount of global memory:                 4096 MBytes (4294967296 bytes)
  ( 5) Multiprocessors, (128) CUDA Cores/MP:       640 CUDA Cores
  GPU Max Clock rate:                                 1124 MHz (1.12 GHz)
  Memory Clock rate:                              1001 Mhz
  Memory Bus Width:                                   128-bit
  L2 Cache Size:                                    2097152 bytes
  Maximum Texture Dimension Size (x,y,z)           1D=(65536), 2D=(65536, 65536), 3D=(4096, 4096, 4096)
  Maximum Layered 1D Texture Size, (num) layers  1      D=(16384), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers    2D=(16384, 16384), 2048 layers
  Total amount of constant memory:                  65536 bytes
  Total amount of shared memory per block:                49152 bytes
  Total number of registers available per block:    65536
  Warp size:                                            32
  Maximum number of threads per multiprocessor:    2048
  Maximum number of threads per block:             1024
  Max dimension size of a thread block (x,y,z):    (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z):    (2147483647, 65535, 65535)
  Maximum memory pitch:                            2147483647 bytes
  Texture alignment:                                512 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):          Yes
  Device PCI Domain ID / Bus ID / location ID:      0 / 1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 8.0, CUDA Runtime Version = 8.0, NumDevs = 1, Device0 = GeForce GTX 950M

Result = PASS

整理：

GeForce GTX 950:
Maximum number of threads per multiprocessor:  2048
Maximum number of threads per block:           1024
registers available per block:   65536个 = 32bits * 65536 = 256kb
(If) registers available per thread:   64  = 64*32bits = 2kb
shared memory per block :   48kb
5 SM, 128 SP:       640 SPs
number of SP per SM:   128
number of block per SM:   2
number of SP per block:   64
number of cycles per block:   16
warp size :   32
number of warp per block:   2
number of thread per warp:   32