/***
*列印顯示卡屬性
*2017-11-8
* ***/
#includeusing namespace std;
int main()
{ int devicecount;
cudagetdevicecount(&devicecount); //returns in *devicecount the number of devices
cout<<"devicecount: "《執行結果如下：
devicecount: 1
name: geforce gtx 1070
totalglobalmem: 8504279040
regsperblock: 65536
warpsize: 32
mempitch: 2147483647
乙個執行緒塊中可使用的最大共享記憶體
devprops.sharedmemperblock: 49152
乙個執行緒塊中可包含的最大執行緒數量
maxthreadsperblock: 1024
多維執行緒塊陣列中每一維可包含的最大執行緒數量
maxthreadsdim[0]: 1024
maxthreadsdim[1]: 1024
maxthreadsdim[2]: 64
乙個執行緒格中每一維可包含的最大執行緒塊數量
maxgridsize[0]: 2147483647
maxgridsize[1]: 65535
maxgridsize[2]: 65535
clockrate: 1784500
totalconstmem: 65536
texturealignment: 512
計算能力：6.1
minor: 1
texturepitchalignment: 32
deviceoverlap: 1
multiprocessorcount: 15
kernelexectimeoutenabled: 1
integrated: 0
canmaphostmemory: 1
computemode: 0
maxtexture1d: 131072
maxtexture1dmipmap: 16384
maxtexture1dlinear: 134217728
maxtexture2d: 0x7fff28079944
maxtexture2dmipmap: 0x7fff2807994c
maxtexture2dlinear: 0x7fff28079954
maxtexture2dgather: 0x7fff28079960
maxtexture3d: 0x7fff28079968
maxtexture3dalt: 0x7fff28079974
maxtexturecubemap: 32768
maxtexture1dlayered: 0x7fff28079984
maxtexture2dlayered: 0x7fff2807998c
maxtexturecubemaplayered: 0x7fff28079998
maxsu***ce1d: 32768
maxsu***ce2d: 0x7fff280799a4
maxsu***ce3d: 0x7fff280799ac
maxsu***ce1dlayered: 0x7fff280799b8
maxsu***ce2dlayered: 0x7fff280799c0
maxsu***cecubemap: 32768
maxsu***cecubemaplayered: 0x7fff280799d0
su***cealignment: 512
concurrentkernels: 1
eccenabled: 0
pcibusid: 2
pcideviceid: 0
pcidomainid: 0
tccdriver: 0
asyncenginecount: 2
unifiedaddressing: 1
memoryclockrate: 4004000
memorybuswidth: 256
l2cachesize: 2097152
maxthreadspermultiprocessor: 2048
streamprioritiessupported: 1
globall1cachesupported: 1
locall1cachesupported: 1
sharedmempermultiprocessor: 98304
regspermultiprocessor: 65536
ismultigpuboard: 0
multigpuboardgroupid: 0
singletodoubleprecisionperfratio: 32
pageablememoryaccess: 0
concurrentmanagedaccess: 1
NVIDIA顯示卡驅動 CUDA相關
顯示卡驅動與cuda版本對應 cuda可以安裝多個版本，在使用時可以進行切換 具體方法 在 usr local 目錄下，有安裝好的多個版本的cuda資料夾，例如cuda 10.0,cuda 9.2,另外還有乙個名為 cuda 的資料夾，是系統讀取cuda設定的資料夾，切換版本就是通過設定cuda 1...
CUDA入門2 獲取顯示卡引數
ifndef cudainput h define cudainput h include cuda runtime.h include device launch parameters.h include typedef enum ci result class input engine endi...
nvidia顯示卡驅動和cuda安裝
1 新增驅動源 sudo add apt repository ppa graphics drivers ppa sudo apt get update 2 安裝nvidia驅動 sudo apt get install nvidia 367 sudo apt get install mesa co...