NASP/cuda-samples
85
0
Fork 0
mirror of https://github.com/NVIDIA/cuda-samples.git synced 2026-10-25 22:28:17 +08:00
Code Issues Packages Projects Releases Wiki Activity

Apply consistent code formatting across the repo. Add clang-format and pre-commit hooks.

Browse Source
This commit is contained in:
Rob Armstrong
2025-03-27 10:30:07 -07:00
parent 2cd58fbc9a
commit ceab6e8bcc
782 changed files with 107230 additions and 106548 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1526
Samples/1_Utilities/bandwidthTest/bandwidthTest.cu
View File

File diff suppressed because it is too large Load Diff

487
Samples/1_Utilities/deviceQuery/deviceQuery.cpp
View File

@@ -32,12 +32,11 @@
#include <cuda_runtime.h>
#include <helper_cuda.h>
#include <iostream>
#include <memory>
#include <string>
int *pArgc = NULL;
int *pArgc = NULL;
char **pArgv = NULL;
#if CUDART_VERSION < 5000
@@ -46,19 +45,16 @@ char **pArgv = NULL;
#include <cuda.h>
// This function wraps the CUDA Driver API into a template function
template <class T>
inline void getCudaAttribute(T *attribute, CUdevice_attribute device_attribute,
int device) {
CUresult error = cuDeviceGetAttribute(attribute, device_attribute, device);
template <class T> inline void getCudaAttribute(T *attribute, CUdevice_attribute device_attribute, int device)
{
CUresult error = cuDeviceGetAttribute(attribute, device_attribute, device);
if (CUDA_SUCCESS != error) {
fprintf(
stderr,
"cuSafeCallNoSync() Driver API error = %04d from file <%s>, line %i.\n",
error, __FILE__, __LINE__);
if (CUDA_SUCCESS != error) {
fprintf(
stderr, "cuSafeCallNoSync() Driver API error = %04d from file <%s>, line %i.\n", error, __FILE__, __LINE__);
exit(EXIT_FAILURE);
}
exit(EXIT_FAILURE);
}
}
#endif /* CUDART_VERSION < 5000 */
@@ -66,278 +62,259 @@ inline void getCudaAttribute(T *attribute, CUdevice_attribute device_attribute,
////////////////////////////////////////////////////////////////////////////////
// Program main
////////////////////////////////////////////////////////////////////////////////
int main(int argc, char **argv) {
pArgc = &argc;
pArgv = argv;
int main(int argc, char **argv)
{
pArgc = &argc;
pArgv = argv;
printf("%s Starting...\n\n", argv[0]);
printf(
" CUDA Device Query (Runtime API) version (CUDART static linking)\n\n");
printf("%s Starting...\n\n", argv[0]);
printf(" CUDA Device Query (Runtime API) version (CUDART static linking)\n\n");
int deviceCount = 0;
cudaError_t error_id = cudaGetDeviceCount(&deviceCount);
int deviceCount = 0;
cudaError_t error_id = cudaGetDeviceCount(&deviceCount);
if (error_id != cudaSuccess) {
printf("cudaGetDeviceCount returned %d\n-> %s\n",
static_cast<int>(error_id), cudaGetErrorString(error_id));
printf("Result = FAIL\n");
exit(EXIT_FAILURE);
}
if (error_id != cudaSuccess) {
printf("cudaGetDeviceCount returned %d\n-> %s\n", static_cast<int>(error_id), cudaGetErrorString(error_id));
printf("Result = FAIL\n");
exit(EXIT_FAILURE);
}
// This function call returns 0 if there are no CUDA capable devices.
if (deviceCount == 0) {
printf("There are no available device(s) that support CUDA\n");
} else {
printf("Detected %d CUDA Capable device(s)\n", deviceCount);
}
// This function call returns 0 if there are no CUDA capable devices.
if (deviceCount == 0) {
printf("There are no available device(s) that support CUDA\n");
}
else {
printf("Detected %d CUDA Capable device(s)\n", deviceCount);
}
int dev, driverVersion = 0, runtimeVersion = 0;
int dev, driverVersion = 0, runtimeVersion = 0;
for (dev = 0; dev < deviceCount; ++dev) {
cudaSetDevice(dev);
cudaDeviceProp deviceProp;
cudaGetDeviceProperties(&deviceProp, dev);
for (dev = 0; dev < deviceCount; ++dev) {
cudaSetDevice(dev);
cudaDeviceProp deviceProp;
cudaGetDeviceProperties(&deviceProp, dev);
printf("\nDevice %d: \"%s\"\n", dev, deviceProp.name);
printf("\nDevice %d: \"%s\"\n", dev, deviceProp.name);
// Console log
cudaDriverGetVersion(&driverVersion);
cudaRuntimeGetVersion(&runtimeVersion);
printf(" CUDA Driver Version / Runtime Version %d.%d / %d.%d\n",
driverVersion / 1000, (driverVersion % 100) / 10,
runtimeVersion / 1000, (runtimeVersion % 100) / 10);
printf(" CUDA Capability Major/Minor version number: %d.%d\n",
deviceProp.major, deviceProp.minor);
// Console log
cudaDriverGetVersion(&driverVersion);
cudaRuntimeGetVersion(&runtimeVersion);
printf(" CUDA Driver Version / Runtime Version %d.%d / %d.%d\n",
driverVersion / 1000,
(driverVersion % 100) / 10,
runtimeVersion / 1000,
(runtimeVersion % 100) / 10);
printf(" CUDA Capability Major/Minor version number: %d.%d\n", deviceProp.major, deviceProp.minor);
char msg[256];
char msg[256];
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
sprintf_s(msg, sizeof(msg),
" Total amount of global memory: %.0f MBytes "
"(%llu bytes)\n",
static_cast<float>(deviceProp.totalGlobalMem / 1048576.0f),
(unsigned long long)deviceProp.totalGlobalMem);
sprintf_s(msg,
sizeof(msg),
" Total amount of global memory: %.0f MBytes "
"(%llu bytes)\n",
static_cast<float>(deviceProp.totalGlobalMem / 1048576.0f),
(unsigned long long)deviceProp.totalGlobalMem);
#else
snprintf(msg, sizeof(msg),
" Total amount of global memory: %.0f MBytes "
"(%llu bytes)\n",
static_cast<float>(deviceProp.totalGlobalMem / 1048576.0f),
(unsigned long long)deviceProp.totalGlobalMem);
snprintf(msg,
sizeof(msg),
" Total amount of global memory: %.0f MBytes "
"(%llu bytes)\n",
static_cast<float>(deviceProp.totalGlobalMem / 1048576.0f),
(unsigned long long)deviceProp.totalGlobalMem);
#endif
printf("%s", msg);
printf("%s", msg);
printf(" (%03d) Multiprocessors, (%03d) CUDA Cores/MP: %d CUDA Cores\n",
deviceProp.multiProcessorCount,
_ConvertSMVer2Cores(deviceProp.major, deviceProp.minor),
_ConvertSMVer2Cores(deviceProp.major, deviceProp.minor) *
deviceProp.multiProcessorCount);
printf(
" GPU Max Clock rate: %.0f MHz (%0.2f "
"GHz)\n",
deviceProp.clockRate * 1e-3f, deviceProp.clockRate * 1e-6f);
printf(" (%03d) Multiprocessors, (%03d) CUDA Cores/MP: %d CUDA Cores\n",
deviceProp.multiProcessorCount,
_ConvertSMVer2Cores(deviceProp.major, deviceProp.minor),
_ConvertSMVer2Cores(deviceProp.major, deviceProp.minor) * deviceProp.multiProcessorCount);
printf(" GPU Max Clock rate: %.0f MHz (%0.2f "
"GHz)\n",
deviceProp.clockRate * 1e-3f,
deviceProp.clockRate * 1e-6f);
#if CUDART_VERSION >= 5000
// This is supported in CUDA 5.0 (runtime API device properties)
printf(" Memory Clock rate: %.0f Mhz\n",
deviceProp.memoryClockRate * 1e-3f);
printf(" Memory Bus Width: %d-bit\n",
deviceProp.memoryBusWidth);
// This is supported in CUDA 5.0 (runtime API device properties)
printf(" Memory Clock rate: %.0f Mhz\n", deviceProp.memoryClockRate * 1e-3f);
printf(" Memory Bus Width: %d-bit\n", deviceProp.memoryBusWidth);
if (deviceProp.l2CacheSize) {
printf(" L2 Cache Size: %d bytes\n",
deviceProp.l2CacheSize);
}
#else
// This only available in CUDA 4.0-4.2 (but these were only exposed in the
// CUDA Driver API)
int memoryClock;
getCudaAttribute<int>(&memoryClock, CU_DEVICE_ATTRIBUTE_MEMORY_CLOCK_RATE,
dev);
printf(" Memory Clock rate: %.0f Mhz\n",
memoryClock * 1e-3f);
int memBusWidth;
getCudaAttribute<int>(&memBusWidth,
CU_DEVICE_ATTRIBUTE_GLOBAL_MEMORY_BUS_WIDTH, dev);
printf(" Memory Bus Width: %d-bit\n",
memBusWidth);
int L2CacheSize;
getCudaAttribute<int>(&L2CacheSize, CU_DEVICE_ATTRIBUTE_L2_CACHE_SIZE, dev);
if (L2CacheSize) {
printf(" L2 Cache Size: %d bytes\n",
L2CacheSize);
}
#endif
printf(
" Maximum Texture Dimension Size (x,y,z) 1D=(%d), 2D=(%d, "
"%d), 3D=(%d, %d, %d)\n",
deviceProp.maxTexture1D, deviceProp.maxTexture2D[0],
deviceProp.maxTexture2D[1], deviceProp.maxTexture3D[0],
deviceProp.maxTexture3D[1], deviceProp.maxTexture3D[2]);
printf(
" Maximum Layered 1D Texture Size, (num) layers 1D=(%d), %d layers\n",
deviceProp.maxTexture1DLayered[0], deviceProp.maxTexture1DLayered[1]);
printf(
" Maximum Layered 2D Texture Size, (num) layers 2D=(%d, %d), %d "
"layers\n",
deviceProp.maxTexture2DLayered[0], deviceProp.maxTexture2DLayered[1],
deviceProp.maxTexture2DLayered[2]);
printf(" Total amount of constant memory: %zu bytes\n",
deviceProp.totalConstMem);
printf(" Total amount of shared memory per block: %zu bytes\n",
deviceProp.sharedMemPerBlock);
printf(" Total shared memory per multiprocessor: %zu bytes\n",
deviceProp.sharedMemPerMultiprocessor);
printf(" Total number of registers available per block: %d\n",
deviceProp.regsPerBlock);
printf(" Warp size: %d\n",
deviceProp.warpSize);
printf(" Maximum number of threads per multiprocessor: %d\n",
deviceProp.maxThreadsPerMultiProcessor);
printf(" Maximum number of threads per block: %d\n",
deviceProp.maxThreadsPerBlock);
printf(" Max dimension size of a thread block (x,y,z): (%d, %d, %d)\n",
deviceProp.maxThreadsDim[0], deviceProp.maxThreadsDim[1],
deviceProp.maxThreadsDim[2]);
printf(" Max dimension size of a grid size (x,y,z): (%d, %d, %d)\n",
deviceProp.maxGridSize[0], deviceProp.maxGridSize[1],
deviceProp.maxGridSize[2]);
printf(" Maximum memory pitch: %zu bytes\n",
deviceProp.memPitch);
printf(" Texture alignment: %zu bytes\n",
deviceProp.textureAlignment);
printf(
" Concurrent copy and kernel execution: %s with %d copy "
"engine(s)\n",
(deviceProp.deviceOverlap ? "Yes" : "No"), deviceProp.asyncEngineCount);
printf(" Run time limit on kernels: %s\n",
deviceProp.kernelExecTimeoutEnabled ? "Yes" : "No");
printf(" Integrated GPU sharing Host Memory: %s\n",
deviceProp.integrated ? "Yes" : "No");
printf(" Support host page-locked memory mapping: %s\n",
deviceProp.canMapHostMemory ? "Yes" : "No");
printf(" Alignment requirement for Surfaces: %s\n",
deviceProp.surfaceAlignment ? "Yes" : "No");
printf(" Device has ECC support: %s\n",
deviceProp.ECCEnabled ? "Enabled" : "Disabled");
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
printf(" CUDA Device Driver Mode (TCC or WDDM): %s\n",
deviceProp.tccDriver ? "TCC (Tesla Compute Cluster Driver)"
: "WDDM (Windows Display Driver Model)");
#endif
printf(" Device supports Unified Addressing (UVA): %s\n",
deviceProp.unifiedAddressing ? "Yes" : "No");
printf(" Device supports Managed Memory: %s\n",
deviceProp.managedMemory ? "Yes" : "No");
printf(" Device supports Compute Preemption: %s\n",
deviceProp.computePreemptionSupported ? "Yes" : "No");
printf(" Supports Cooperative Kernel Launch: %s\n",
deviceProp.cooperativeLaunch ? "Yes" : "No");
printf(" Supports MultiDevice Co-op Kernel Launch: %s\n",
deviceProp.cooperativeMultiDeviceLaunch ? "Yes" : "No");
printf(" Device PCI Domain ID / Bus ID / location ID: %d / %d / %d\n",
deviceProp.pciDomainID, deviceProp.pciBusID, deviceProp.pciDeviceID);
const char *sComputeMode[] = {
"Default (multiple host threads can use ::cudaSetDevice() with device "
"simultaneously)",
"Exclusive (only one host thread in one process is able to use "
"::cudaSetDevice() with this device)",
"Prohibited (no host thread can use ::cudaSetDevice() with this "
"device)",
"Exclusive Process (many threads in one process is able to use "
"::cudaSetDevice() with this device)",
"Unknown", NULL};
printf(" Compute Mode:\n");
printf(" < %s >\n", sComputeMode[deviceProp.computeMode]);
}
// If there are 2 or more GPUs, query to determine whether RDMA is supported
if (deviceCount >= 2) {
cudaDeviceProp prop[64];
int gpuid[64]; // we want to find the first two GPUs that can support P2P
int gpu_p2p_count = 0;
for (int i = 0; i < deviceCount; i++) {
checkCudaErrors(cudaGetDeviceProperties(&prop[i], i));
// Only boards based on Fermi or later can support P2P
if ((prop[i].major >= 2)
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
// on Windows (64-bit), the Tesla Compute Cluster driver for windows
// must be enabled to support this
&& prop[i].tccDriver
#endif
) {
// This is an array of P2P capable GPUs
gpuid[gpu_p2p_count++] = i;
}
}
// Show all the combinations of support P2P GPUs
int can_access_peer;
if (gpu_p2p_count >= 2) {
for (int i = 0; i < gpu_p2p_count; i++) {
for (int j = 0; j < gpu_p2p_count; j++) {
if (gpuid[i] == gpuid[j]) {
continue;
}
checkCudaErrors(
cudaDeviceCanAccessPeer(&can_access_peer, gpuid[i], gpuid[j]));
printf("> Peer access from %s (GPU%d) -> %s (GPU%d) : %s\n",
prop[gpuid[i]].name, gpuid[i], prop[gpuid[j]].name, gpuid[j],
can_access_peer ? "Yes" : "No");
if (deviceProp.l2CacheSize) {
printf(" L2 Cache Size: %d bytes\n", deviceProp.l2CacheSize);
}
}
#else
// This only available in CUDA 4.0-4.2 (but these were only exposed in the
// CUDA Driver API)
int memoryClock;
getCudaAttribute<int>(&memoryClock, CU_DEVICE_ATTRIBUTE_MEMORY_CLOCK_RATE, dev);
printf(" Memory Clock rate: %.0f Mhz\n", memoryClock * 1e-3f);
int memBusWidth;
getCudaAttribute<int>(&memBusWidth, CU_DEVICE_ATTRIBUTE_GLOBAL_MEMORY_BUS_WIDTH, dev);
printf(" Memory Bus Width: %d-bit\n", memBusWidth);
int L2CacheSize;
getCudaAttribute<int>(&L2CacheSize, CU_DEVICE_ATTRIBUTE_L2_CACHE_SIZE, dev);
if (L2CacheSize) {
printf(" L2 Cache Size: %d bytes\n", L2CacheSize);
}
#endif
printf(" Maximum Texture Dimension Size (x,y,z) 1D=(%d), 2D=(%d, "
"%d), 3D=(%d, %d, %d)\n",
deviceProp.maxTexture1D,
deviceProp.maxTexture2D[0],
deviceProp.maxTexture2D[1],
deviceProp.maxTexture3D[0],
deviceProp.maxTexture3D[1],
deviceProp.maxTexture3D[2]);
printf(" Maximum Layered 1D Texture Size, (num) layers 1D=(%d), %d layers\n",
deviceProp.maxTexture1DLayered[0],
deviceProp.maxTexture1DLayered[1]);
printf(" Maximum Layered 2D Texture Size, (num) layers 2D=(%d, %d), %d "
"layers\n",
deviceProp.maxTexture2DLayered[0],
deviceProp.maxTexture2DLayered[1],
deviceProp.maxTexture2DLayered[2]);
printf(" Total amount of constant memory: %zu bytes\n", deviceProp.totalConstMem);
printf(" Total amount of shared memory per block: %zu bytes\n", deviceProp.sharedMemPerBlock);
printf(" Total shared memory per multiprocessor: %zu bytes\n", deviceProp.sharedMemPerMultiprocessor);
printf(" Total number of registers available per block: %d\n", deviceProp.regsPerBlock);
printf(" Warp size: %d\n", deviceProp.warpSize);
printf(" Maximum number of threads per multiprocessor: %d\n", deviceProp.maxThreadsPerMultiProcessor);
printf(" Maximum number of threads per block: %d\n", deviceProp.maxThreadsPerBlock);
printf(" Max dimension size of a thread block (x,y,z): (%d, %d, %d)\n",
deviceProp.maxThreadsDim[0],
deviceProp.maxThreadsDim[1],
deviceProp.maxThreadsDim[2]);
printf(" Max dimension size of a grid size (x,y,z): (%d, %d, %d)\n",
deviceProp.maxGridSize[0],
deviceProp.maxGridSize[1],
deviceProp.maxGridSize[2]);
printf(" Maximum memory pitch: %zu bytes\n", deviceProp.memPitch);
printf(" Texture alignment: %zu bytes\n", deviceProp.textureAlignment);
printf(" Concurrent copy and kernel execution: %s with %d copy "
"engine(s)\n",
(deviceProp.deviceOverlap ? "Yes" : "No"),
deviceProp.asyncEngineCount);
printf(" Run time limit on kernels: %s\n",
deviceProp.kernelExecTimeoutEnabled ? "Yes" : "No");
printf(" Integrated GPU sharing Host Memory: %s\n", deviceProp.integrated ? "Yes" : "No");
printf(" Support host page-locked memory mapping: %s\n", deviceProp.canMapHostMemory ? "Yes" : "No");
printf(" Alignment requirement for Surfaces: %s\n", deviceProp.surfaceAlignment ? "Yes" : "No");
printf(" Device has ECC support: %s\n", deviceProp.ECCEnabled ? "Enabled" : "Disabled");
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
printf(" CUDA Device Driver Mode (TCC or WDDM): %s\n",
deviceProp.tccDriver ? "TCC (Tesla Compute Cluster Driver)" : "WDDM (Windows Display Driver Model)");
#endif
printf(" Device supports Unified Addressing (UVA): %s\n", deviceProp.unifiedAddressing ? "Yes" : "No");
printf(" Device supports Managed Memory: %s\n", deviceProp.managedMemory ? "Yes" : "No");
printf(" Device supports Compute Preemption: %s\n",
deviceProp.computePreemptionSupported ? "Yes" : "No");
printf(" Supports Cooperative Kernel Launch: %s\n", deviceProp.cooperativeLaunch ? "Yes" : "No");
printf(" Supports MultiDevice Co-op Kernel Launch: %s\n",
deviceProp.cooperativeMultiDeviceLaunch ? "Yes" : "No");
printf(" Device PCI Domain ID / Bus ID / location ID: %d / %d / %d\n",
deviceProp.pciDomainID,
deviceProp.pciBusID,
deviceProp.pciDeviceID);
const char *sComputeMode[] = {"Default (multiple host threads can use ::cudaSetDevice() with device "
"simultaneously)",
"Exclusive (only one host thread in one process is able to use "
"::cudaSetDevice() with this device)",
"Prohibited (no host thread can use ::cudaSetDevice() with this "
"device)",
"Exclusive Process (many threads in one process is able to use "
"::cudaSetDevice() with this device)",
"Unknown",
NULL};
printf(" Compute Mode:\n");
printf(" < %s >\n", sComputeMode[deviceProp.computeMode]);
}
}
// csv masterlog info
// *****************************
// exe and CUDA driver name
printf("\n");
std::string sProfileString = "deviceQuery, CUDA Driver = CUDART";
char cTemp[16];
// If there are 2 or more GPUs, query to determine whether RDMA is supported
if (deviceCount >= 2) {
cudaDeviceProp prop[64];
int gpuid[64]; // we want to find the first two GPUs that can support P2P
int gpu_p2p_count = 0;
// driver version
sProfileString += ", CUDA Driver Version = ";
for (int i = 0; i < deviceCount; i++) {
checkCudaErrors(cudaGetDeviceProperties(&prop[i], i));
// Only boards based on Fermi or later can support P2P
if ((prop[i].major >= 2)
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
sprintf_s(cTemp, 10, "%d.%d", driverVersion / 1000,
(driverVersion % 100) / 10);
#else
snprintf(cTemp, sizeof(cTemp), "%d.%d", driverVersion / 1000,
(driverVersion % 100) / 10);
// on Windows (64-bit), the Tesla Compute Cluster driver for windows
// must be enabled to support this
&& prop[i].tccDriver
#endif
sProfileString += cTemp;
) {
// This is an array of P2P capable GPUs
gpuid[gpu_p2p_count++] = i;
}
}
// Runtime version
sProfileString += ", CUDA Runtime Version = ";
// Show all the combinations of support P2P GPUs
int can_access_peer;
if (gpu_p2p_count >= 2) {
for (int i = 0; i < gpu_p2p_count; i++) {
for (int j = 0; j < gpu_p2p_count; j++) {
if (gpuid[i] == gpuid[j]) {
continue;
}
checkCudaErrors(cudaDeviceCanAccessPeer(&can_access_peer, gpuid[i], gpuid[j]));
printf("> Peer access from %s (GPU%d) -> %s (GPU%d) : %s\n",
prop[gpuid[i]].name,
gpuid[i],
prop[gpuid[j]].name,
gpuid[j],
can_access_peer ? "Yes" : "No");
}
}
}
}
// csv masterlog info
// *****************************
// exe and CUDA driver name
printf("\n");
std::string sProfileString = "deviceQuery, CUDA Driver = CUDART";
char cTemp[16];
// driver version
sProfileString += ", CUDA Driver Version = ";
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
sprintf_s(cTemp, 10, "%d.%d", runtimeVersion / 1000,
(runtimeVersion % 100) / 10);
sprintf_s(cTemp, 10, "%d.%d", driverVersion / 1000, (driverVersion % 100) / 10);
#else
snprintf(cTemp, sizeof(cTemp), "%d.%d", runtimeVersion / 1000,
(runtimeVersion % 100) / 10);
snprintf(cTemp, sizeof(cTemp), "%d.%d", driverVersion / 1000, (driverVersion % 100) / 10);
#endif
sProfileString += cTemp;
sProfileString += cTemp;
// Device count
sProfileString += ", NumDevs = ";
// Runtime version
sProfileString += ", CUDA Runtime Version = ";
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
sprintf_s(cTemp, 10, "%d", deviceCount);
sprintf_s(cTemp, 10, "%d.%d", runtimeVersion / 1000, (runtimeVersion % 100) / 10);
#else
snprintf(cTemp, sizeof(cTemp), "%d", deviceCount);
snprintf(cTemp, sizeof(cTemp), "%d.%d", runtimeVersion / 1000, (runtimeVersion % 100) / 10);
#endif
sProfileString += cTemp;
sProfileString += "\n";
printf("%s", sProfileString.c_str());
sProfileString += cTemp;
printf("Result = PASS\n");
// Device count
sProfileString += ", NumDevs = ";
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
sprintf_s(cTemp, 10, "%d", deviceCount);
#else
snprintf(cTemp, sizeof(cTemp), "%d", deviceCount);
#endif
sProfileString += cTemp;
sProfileString += "\n";
printf("%s", sProfileString.c_str());
// finish
exit(EXIT_SUCCESS);
printf("Result = PASS\n");
// finish
exit(EXIT_SUCCESS);
}

593
Samples/1_Utilities/deviceQueryDrv/deviceQueryDrv.cpp
View File

@@ -30,358 +30,295 @@
*/
// includes, system
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <cuda.h>
#include <helper_cuda_drvapi.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
////////////////////////////////////////////////////////////////////////////////
// Program main
////////////////////////////////////////////////////////////////////////////////
int main(int argc, char **argv) {
CUdevice dev;
int major = 0, minor = 0;
int deviceCount = 0;
char deviceName[256];
int main(int argc, char **argv)
{
CUdevice dev;
int major = 0, minor = 0;
int deviceCount = 0;
char deviceName[256];
printf("%s Starting...\n\n", argv[0]);
printf("%s Starting...\n\n", argv[0]);
// note your project will need to link with cuda.lib files on windows
printf("CUDA Device Query (Driver API) statically linked version \n");
// note your project will need to link with cuda.lib files on windows
printf("CUDA Device Query (Driver API) statically linked version \n");
checkCudaErrors(cuInit(0));
checkCudaErrors(cuInit(0));
checkCudaErrors(cuDeviceGetCount(&deviceCount));
checkCudaErrors(cuDeviceGetCount(&deviceCount));
// This function call returns 0 if there are no CUDA capable devices.
if (deviceCount == 0) {
printf("There are no available device(s) that support CUDA\n");
} else {
printf("Detected %d CUDA Capable device(s)\n", deviceCount);
}
for (dev = 0; dev < deviceCount; ++dev) {
checkCudaErrors(cuDeviceGetAttribute(
&major, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR, dev));
checkCudaErrors(cuDeviceGetAttribute(
&minor, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR, dev));
checkCudaErrors(cuDeviceGetName(deviceName, 256, dev));
printf("\nDevice %d: \"%s\"\n", dev, deviceName);
int driverVersion = 0;
checkCudaErrors(cuDriverGetVersion(&driverVersion));
printf(" CUDA Driver Version: %d.%d\n",
driverVersion / 1000, (driverVersion % 100) / 10);
printf(" CUDA Capability Major/Minor version number: %d.%d\n", major,
minor);
size_t totalGlobalMem;
checkCudaErrors(cuDeviceTotalMem(&totalGlobalMem, dev));
char msg[256];
SPRINTF(msg,
" Total amount of global memory: %.0f MBytes "
"(%llu bytes)\n",
(float)totalGlobalMem / 1048576.0f,
(unsigned long long)totalGlobalMem);
printf("%s", msg);
int multiProcessorCount;
getCudaAttribute<int>(&multiProcessorCount,
CU_DEVICE_ATTRIBUTE_MULTIPROCESSOR_COUNT, dev);
printf(" (%2d) Multiprocessors, (%3d) CUDA Cores/MP: %d CUDA Cores\n",
multiProcessorCount, _ConvertSMVer2CoresDRV(major, minor),
_ConvertSMVer2CoresDRV(major, minor) * multiProcessorCount);
int clockRate;
getCudaAttribute<int>(&clockRate, CU_DEVICE_ATTRIBUTE_CLOCK_RATE, dev);
printf(
" GPU Max Clock rate: %.0f MHz (%0.2f "
"GHz)\n",
clockRate * 1e-3f, clockRate * 1e-6f);
int memoryClock;
getCudaAttribute<int>(&memoryClock, CU_DEVICE_ATTRIBUTE_MEMORY_CLOCK_RATE,
dev);
printf(" Memory Clock rate: %.0f Mhz\n",
memoryClock * 1e-3f);
int memBusWidth;
getCudaAttribute<int>(&memBusWidth,
CU_DEVICE_ATTRIBUTE_GLOBAL_MEMORY_BUS_WIDTH, dev);
printf(" Memory Bus Width: %d-bit\n",
memBusWidth);
int L2CacheSize;
getCudaAttribute<int>(&L2CacheSize, CU_DEVICE_ATTRIBUTE_L2_CACHE_SIZE, dev);
if (L2CacheSize) {
printf(" L2 Cache Size: %d bytes\n",
L2CacheSize);
// This function call returns 0 if there are no CUDA capable devices.
if (deviceCount == 0) {
printf("There are no available device(s) that support CUDA\n");
}
else {
printf("Detected %d CUDA Capable device(s)\n", deviceCount);
}
int maxTex1D, maxTex2D[2], maxTex3D[3];
getCudaAttribute<int>(&maxTex1D,
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE1D_WIDTH, dev);
getCudaAttribute<int>(&maxTex2D[0],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_WIDTH, dev);
getCudaAttribute<int>(&maxTex2D[1],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_HEIGHT, dev);
getCudaAttribute<int>(&maxTex3D[0],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE3D_WIDTH, dev);
getCudaAttribute<int>(&maxTex3D[1],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE3D_HEIGHT, dev);
getCudaAttribute<int>(&maxTex3D[2],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE3D_DEPTH, dev);
printf(
" Max Texture Dimension Sizes 1D=(%d) 2D=(%d, %d) "
"3D=(%d, %d, %d)\n",
maxTex1D, maxTex2D[0], maxTex2D[1], maxTex3D[0], maxTex3D[1],
maxTex3D[2]);
for (dev = 0; dev < deviceCount; ++dev) {
checkCudaErrors(cuDeviceGetAttribute(&major, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR, dev));
checkCudaErrors(cuDeviceGetAttribute(&minor, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR, dev));
int maxTex1DLayered[2];
getCudaAttribute<int>(&maxTex1DLayered[0],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE1D_LAYERED_WIDTH,
dev);
getCudaAttribute<int>(&maxTex1DLayered[1],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE1D_LAYERED_LAYERS,
dev);
printf(
" Maximum Layered 1D Texture Size, (num) layers 1D=(%d), %d layers\n",
maxTex1DLayered[0], maxTex1DLayered[1]);
checkCudaErrors(cuDeviceGetName(deviceName, 256, dev));
int maxTex2DLayered[3];
getCudaAttribute<int>(&maxTex2DLayered[0],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_LAYERED_WIDTH,
dev);
getCudaAttribute<int>(&maxTex2DLayered[1],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_LAYERED_HEIGHT,
dev);
getCudaAttribute<int>(&maxTex2DLayered[2],
CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_LAYERED_LAYERS,
dev);
printf(
" Maximum Layered 2D Texture Size, (num) layers 2D=(%d, %d), %d "
"layers\n",
maxTex2DLayered[0], maxTex2DLayered[1], maxTex2DLayered[2]);
printf("\nDevice %d: \"%s\"\n", dev, deviceName);
int totalConstantMemory;
getCudaAttribute<int>(&totalConstantMemory,
CU_DEVICE_ATTRIBUTE_TOTAL_CONSTANT_MEMORY, dev);
printf(" Total amount of constant memory: %u bytes\n",
totalConstantMemory);
int sharedMemPerBlock;
getCudaAttribute<int>(&sharedMemPerBlock,
CU_DEVICE_ATTRIBUTE_MAX_SHARED_MEMORY_PER_BLOCK, dev);
printf(" Total amount of shared memory per block: %u bytes\n",
sharedMemPerBlock);
int regsPerBlock;
getCudaAttribute<int>(&regsPerBlock,
CU_DEVICE_ATTRIBUTE_MAX_REGISTERS_PER_BLOCK, dev);
printf(" Total number of registers available per block: %d\n",
regsPerBlock);
int warpSize;
getCudaAttribute<int>(&warpSize, CU_DEVICE_ATTRIBUTE_WARP_SIZE, dev);
printf(" Warp size: %d\n", warpSize);
int maxThreadsPerMultiProcessor;
getCudaAttribute<int>(&maxThreadsPerMultiProcessor,
CU_DEVICE_ATTRIBUTE_MAX_THREADS_PER_MULTIPROCESSOR,
dev);
printf(" Maximum number of threads per multiprocessor: %d\n",
maxThreadsPerMultiProcessor);
int maxThreadsPerBlock;
getCudaAttribute<int>(&maxThreadsPerBlock,
CU_DEVICE_ATTRIBUTE_MAX_THREADS_PER_BLOCK, dev);
printf(" Maximum number of threads per block: %d\n",
maxThreadsPerBlock);
int driverVersion = 0;
checkCudaErrors(cuDriverGetVersion(&driverVersion));
printf(" CUDA Driver Version: %d.%d\n",
driverVersion / 1000,
(driverVersion % 100) / 10);
printf(" CUDA Capability Major/Minor version number: %d.%d\n", major, minor);
int blockDim[3];
getCudaAttribute<int>(&blockDim[0], CU_DEVICE_ATTRIBUTE_MAX_BLOCK_DIM_X,
dev);
getCudaAttribute<int>(&blockDim[1], CU_DEVICE_ATTRIBUTE_MAX_BLOCK_DIM_Y,
dev);
getCudaAttribute<int>(&blockDim[2], CU_DEVICE_ATTRIBUTE_MAX_BLOCK_DIM_Z,
dev);
printf(" Max dimension size of a thread block (x,y,z): (%d, %d, %d)\n",
blockDim[0], blockDim[1], blockDim[2]);
int gridDim[3];
getCudaAttribute<int>(&gridDim[0], CU_DEVICE_ATTRIBUTE_MAX_GRID_DIM_X, dev);
getCudaAttribute<int>(&gridDim[1], CU_DEVICE_ATTRIBUTE_MAX_GRID_DIM_Y, dev);
getCudaAttribute<int>(&gridDim[2], CU_DEVICE_ATTRIBUTE_MAX_GRID_DIM_Z, dev);
printf(" Max dimension size of a grid size (x,y,z): (%d, %d, %d)\n",
gridDim[0], gridDim[1], gridDim[2]);
size_t totalGlobalMem;
checkCudaErrors(cuDeviceTotalMem(&totalGlobalMem, dev));
int textureAlign;
getCudaAttribute<int>(&textureAlign, CU_DEVICE_ATTRIBUTE_TEXTURE_ALIGNMENT,
dev);
printf(" Texture alignment: %u bytes\n",
textureAlign);
char msg[256];
SPRINTF(msg,
" Total amount of global memory: %.0f MBytes "
"(%llu bytes)\n",
(float)totalGlobalMem / 1048576.0f,
(unsigned long long)totalGlobalMem);
printf("%s", msg);
int memPitch;
getCudaAttribute<int>(&memPitch, CU_DEVICE_ATTRIBUTE_MAX_PITCH, dev);
printf(" Maximum memory pitch: %u bytes\n",
memPitch);
int multiProcessorCount;
getCudaAttribute<int>(&multiProcessorCount, CU_DEVICE_ATTRIBUTE_MULTIPROCESSOR_COUNT, dev);
int gpuOverlap;
getCudaAttribute<int>(&gpuOverlap, CU_DEVICE_ATTRIBUTE_GPU_OVERLAP, dev);
printf(" (%2d) Multiprocessors, (%3d) CUDA Cores/MP: %d CUDA Cores\n",
multiProcessorCount,
_ConvertSMVer2CoresDRV(major, minor),
_ConvertSMVer2CoresDRV(major, minor) * multiProcessorCount);
int asyncEngineCount;
getCudaAttribute<int>(&asyncEngineCount,
CU_DEVICE_ATTRIBUTE_ASYNC_ENGINE_COUNT, dev);
printf(
" Concurrent copy and kernel execution: %s with %d copy "
"engine(s)\n",
(gpuOverlap ? "Yes" : "No"), asyncEngineCount);
int clockRate;
getCudaAttribute<int>(&clockRate, CU_DEVICE_ATTRIBUTE_CLOCK_RATE, dev);
printf(" GPU Max Clock rate: %.0f MHz (%0.2f "
"GHz)\n",
clockRate * 1e-3f,
clockRate * 1e-6f);
int memoryClock;
getCudaAttribute<int>(&memoryClock, CU_DEVICE_ATTRIBUTE_MEMORY_CLOCK_RATE, dev);
printf(" Memory Clock rate: %.0f Mhz\n", memoryClock * 1e-3f);
int memBusWidth;
getCudaAttribute<int>(&memBusWidth, CU_DEVICE_ATTRIBUTE_GLOBAL_MEMORY_BUS_WIDTH, dev);
printf(" Memory Bus Width: %d-bit\n", memBusWidth);
int L2CacheSize;
getCudaAttribute<int>(&L2CacheSize, CU_DEVICE_ATTRIBUTE_L2_CACHE_SIZE, dev);
int kernelExecTimeoutEnabled;
getCudaAttribute<int>(&kernelExecTimeoutEnabled,
CU_DEVICE_ATTRIBUTE_KERNEL_EXEC_TIMEOUT, dev);
printf(" Run time limit on kernels: %s\n",
kernelExecTimeoutEnabled ? "Yes" : "No");
int integrated;
getCudaAttribute<int>(&integrated, CU_DEVICE_ATTRIBUTE_INTEGRATED, dev);
printf(" Integrated GPU sharing Host Memory: %s\n",
integrated ? "Yes" : "No");
int canMapHostMemory;
getCudaAttribute<int>(&canMapHostMemory,
CU_DEVICE_ATTRIBUTE_CAN_MAP_HOST_MEMORY, dev);
printf(" Support host page-locked memory mapping: %s\n",
canMapHostMemory ? "Yes" : "No");
int concurrentKernels;
getCudaAttribute<int>(&concurrentKernels,
CU_DEVICE_ATTRIBUTE_CONCURRENT_KERNELS, dev);
printf(" Concurrent kernel execution: %s\n",
concurrentKernels ? "Yes" : "No");
int surfaceAlignment;
getCudaAttribute<int>(&surfaceAlignment,
CU_DEVICE_ATTRIBUTE_SURFACE_ALIGNMENT, dev);
printf(" Alignment requirement for Surfaces: %s\n",
surfaceAlignment ? "Yes" : "No");
int eccEnabled;
getCudaAttribute<int>(&eccEnabled, CU_DEVICE_ATTRIBUTE_ECC_ENABLED, dev);
printf(" Device has ECC support: %s\n",
eccEnabled ? "Enabled" : "Disabled");
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
int tccDriver;
getCudaAttribute<int>(&tccDriver, CU_DEVICE_ATTRIBUTE_TCC_DRIVER, dev);
printf(" CUDA Device Driver Mode (TCC or WDDM): %s\n",
tccDriver ? "TCC (Tesla Compute Cluster Driver)"
: "WDDM (Windows Display Driver Model)");
#endif
int unifiedAddressing;
getCudaAttribute<int>(&unifiedAddressing,
CU_DEVICE_ATTRIBUTE_UNIFIED_ADDRESSING, dev);
printf(" Device supports Unified Addressing (UVA): %s\n",
unifiedAddressing ? "Yes" : "No");
int managedMemory;
getCudaAttribute<int>(&managedMemory, CU_DEVICE_ATTRIBUTE_MANAGED_MEMORY,
dev);
printf(" Device supports Managed Memory: %s\n",
managedMemory ? "Yes" : "No");
int computePreemption;
getCudaAttribute<int>(&computePreemption,
CU_DEVICE_ATTRIBUTE_COMPUTE_PREEMPTION_SUPPORTED,
dev);
printf(" Device supports Compute Preemption: %s\n",
computePreemption ? "Yes" : "No");
int cooperativeLaunch;
getCudaAttribute<int>(&cooperativeLaunch,
CU_DEVICE_ATTRIBUTE_COOPERATIVE_LAUNCH, dev);
printf(" Supports Cooperative Kernel Launch: %s\n",
cooperativeLaunch ? "Yes" : "No");
int cooperativeMultiDevLaunch;
getCudaAttribute<int>(&cooperativeMultiDevLaunch,
CU_DEVICE_ATTRIBUTE_COOPERATIVE_MULTI_DEVICE_LAUNCH,
dev);
printf(" Supports MultiDevice Co-op Kernel Launch: %s\n",
cooperativeMultiDevLaunch ? "Yes" : "No");
int pciDomainID, pciBusID, pciDeviceID;
getCudaAttribute<int>(&pciDomainID, CU_DEVICE_ATTRIBUTE_PCI_DOMAIN_ID, dev);
getCudaAttribute<int>(&pciBusID, CU_DEVICE_ATTRIBUTE_PCI_BUS_ID, dev);
getCudaAttribute<int>(&pciDeviceID, CU_DEVICE_ATTRIBUTE_PCI_DEVICE_ID, dev);
printf(" Device PCI Domain ID / Bus ID / location ID: %d / %d / %d\n",
pciDomainID, pciBusID, pciDeviceID);
const char *sComputeMode[] = {
"Default (multiple host threads can use ::cudaSetDevice() with device "
"simultaneously)",
"Exclusive (only one host thread in one process is able to use "
"::cudaSetDevice() with this device)",
"Prohibited (no host thread can use ::cudaSetDevice() with this "
"device)",
"Exclusive Process (many threads in one process is able to use "
"::cudaSetDevice() with this device)",
"Unknown", NULL};
int computeMode;
getCudaAttribute<int>(&computeMode, CU_DEVICE_ATTRIBUTE_COMPUTE_MODE, dev);
printf(" Compute Mode:\n");
printf(" < %s >\n", sComputeMode[computeMode]);
}
// If there are 2 or more GPUs, query to determine whether RDMA is supported
if (deviceCount >= 2) {
int gpuid[64]; // we want to find the first two GPUs that can support P2P
int gpu_p2p_count = 0;
int tccDriver = 0;
for (int i = 0; i < deviceCount; i++) {
checkCudaErrors(cuDeviceGetAttribute(
&major, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR, i));
checkCudaErrors(cuDeviceGetAttribute(
&minor, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR, i));
getCudaAttribute<int>(&tccDriver, CU_DEVICE_ATTRIBUTE_TCC_DRIVER, i);
// Only boards based on Fermi or later can support P2P
if ((major >= 2)
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
// on Windows (64-bit), the Tesla Compute Cluster driver for windows
// must be enabled to support this
&& tccDriver
#endif
) {
// This is an array of P2P capable GPUs
gpuid[gpu_p2p_count++] = i;
}
}
// Show all the combinations of support P2P GPUs
int can_access_peer;
char deviceName0[256], deviceName1[256];
if (gpu_p2p_count >= 2) {
for (int i = 0; i < gpu_p2p_count; i++) {
for (int j = 0; j < gpu_p2p_count; j++) {
if (gpuid[i] == gpuid[j]) {
continue;
}
checkCudaErrors(
cuDeviceCanAccessPeer(&can_access_peer, gpuid[i], gpuid[j]));
checkCudaErrors(cuDeviceGetName(deviceName0, 256, gpuid[i]));
checkCudaErrors(cuDeviceGetName(deviceName1, 256, gpuid[j]));
printf(
"> Peer-to-Peer (P2P) access from %s (GPU%d) -> %s (GPU%d) : "
"%s\n",
deviceName0, gpuid[i], deviceName1, gpuid[j],
can_access_peer ? "Yes" : "No");
if (L2CacheSize) {
printf(" L2 Cache Size: %d bytes\n", L2CacheSize);
}
}
int maxTex1D, maxTex2D[2], maxTex3D[3];
getCudaAttribute<int>(&maxTex1D, CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE1D_WIDTH, dev);
getCudaAttribute<int>(&maxTex2D[0], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_WIDTH, dev);
getCudaAttribute<int>(&maxTex2D[1], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_HEIGHT, dev);
getCudaAttribute<int>(&maxTex3D[0], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE3D_WIDTH, dev);
getCudaAttribute<int>(&maxTex3D[1], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE3D_HEIGHT, dev);
getCudaAttribute<int>(&maxTex3D[2], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE3D_DEPTH, dev);
printf(" Max Texture Dimension Sizes 1D=(%d) 2D=(%d, %d) "
"3D=(%d, %d, %d)\n",
maxTex1D,
maxTex2D[0],
maxTex2D[1],
maxTex3D[0],
maxTex3D[1],
maxTex3D[2]);
int maxTex1DLayered[2];
getCudaAttribute<int>(&maxTex1DLayered[0], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE1D_LAYERED_WIDTH, dev);
getCudaAttribute<int>(&maxTex1DLayered[1], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE1D_LAYERED_LAYERS, dev);
printf(" Maximum Layered 1D Texture Size, (num) layers 1D=(%d), %d layers\n",
maxTex1DLayered[0],
maxTex1DLayered[1]);
int maxTex2DLayered[3];
getCudaAttribute<int>(&maxTex2DLayered[0], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_LAYERED_WIDTH, dev);
getCudaAttribute<int>(&maxTex2DLayered[1], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_LAYERED_HEIGHT, dev);
getCudaAttribute<int>(&maxTex2DLayered[2], CU_DEVICE_ATTRIBUTE_MAXIMUM_TEXTURE2D_LAYERED_LAYERS, dev);
printf(" Maximum Layered 2D Texture Size, (num) layers 2D=(%d, %d), %d "
"layers\n",
maxTex2DLayered[0],
maxTex2DLayered[1],
maxTex2DLayered[2]);
int totalConstantMemory;
getCudaAttribute<int>(&totalConstantMemory, CU_DEVICE_ATTRIBUTE_TOTAL_CONSTANT_MEMORY, dev);
printf(" Total amount of constant memory: %u bytes\n", totalConstantMemory);
int sharedMemPerBlock;
getCudaAttribute<int>(&sharedMemPerBlock, CU_DEVICE_ATTRIBUTE_MAX_SHARED_MEMORY_PER_BLOCK, dev);
printf(" Total amount of shared memory per block: %u bytes\n", sharedMemPerBlock);
int regsPerBlock;
getCudaAttribute<int>(&regsPerBlock, CU_DEVICE_ATTRIBUTE_MAX_REGISTERS_PER_BLOCK, dev);
printf(" Total number of registers available per block: %d\n", regsPerBlock);
int warpSize;
getCudaAttribute<int>(&warpSize, CU_DEVICE_ATTRIBUTE_WARP_SIZE, dev);
printf(" Warp size: %d\n", warpSize);
int maxThreadsPerMultiProcessor;
getCudaAttribute<int>(&maxThreadsPerMultiProcessor, CU_DEVICE_ATTRIBUTE_MAX_THREADS_PER_MULTIPROCESSOR, dev);
printf(" Maximum number of threads per multiprocessor: %d\n", maxThreadsPerMultiProcessor);
int maxThreadsPerBlock;
getCudaAttribute<int>(&maxThreadsPerBlock, CU_DEVICE_ATTRIBUTE_MAX_THREADS_PER_BLOCK, dev);
printf(" Maximum number of threads per block: %d\n", maxThreadsPerBlock);
int blockDim[3];
getCudaAttribute<int>(&blockDim[0], CU_DEVICE_ATTRIBUTE_MAX_BLOCK_DIM_X, dev);
getCudaAttribute<int>(&blockDim[1], CU_DEVICE_ATTRIBUTE_MAX_BLOCK_DIM_Y, dev);
getCudaAttribute<int>(&blockDim[2], CU_DEVICE_ATTRIBUTE_MAX_BLOCK_DIM_Z, dev);
printf(" Max dimension size of a thread block (x,y,z): (%d, %d, %d)\n", blockDim[0], blockDim[1], blockDim[2]);
int gridDim[3];
getCudaAttribute<int>(&gridDim[0], CU_DEVICE_ATTRIBUTE_MAX_GRID_DIM_X, dev);
getCudaAttribute<int>(&gridDim[1], CU_DEVICE_ATTRIBUTE_MAX_GRID_DIM_Y, dev);
getCudaAttribute<int>(&gridDim[2], CU_DEVICE_ATTRIBUTE_MAX_GRID_DIM_Z, dev);
printf(" Max dimension size of a grid size (x,y,z): (%d, %d, %d)\n", gridDim[0], gridDim[1], gridDim[2]);
int textureAlign;
getCudaAttribute<int>(&textureAlign, CU_DEVICE_ATTRIBUTE_TEXTURE_ALIGNMENT, dev);
printf(" Texture alignment: %u bytes\n", textureAlign);
int memPitch;
getCudaAttribute<int>(&memPitch, CU_DEVICE_ATTRIBUTE_MAX_PITCH, dev);
printf(" Maximum memory pitch: %u bytes\n", memPitch);
int gpuOverlap;
getCudaAttribute<int>(&gpuOverlap, CU_DEVICE_ATTRIBUTE_GPU_OVERLAP, dev);
int asyncEngineCount;
getCudaAttribute<int>(&asyncEngineCount, CU_DEVICE_ATTRIBUTE_ASYNC_ENGINE_COUNT, dev);
printf(" Concurrent copy and kernel execution: %s with %d copy "
"engine(s)\n",
(gpuOverlap ? "Yes" : "No"),
asyncEngineCount);
int kernelExecTimeoutEnabled;
getCudaAttribute<int>(&kernelExecTimeoutEnabled, CU_DEVICE_ATTRIBUTE_KERNEL_EXEC_TIMEOUT, dev);
printf(" Run time limit on kernels: %s\n", kernelExecTimeoutEnabled ? "Yes" : "No");
int integrated;
getCudaAttribute<int>(&integrated, CU_DEVICE_ATTRIBUTE_INTEGRATED, dev);
printf(" Integrated GPU sharing Host Memory: %s\n", integrated ? "Yes" : "No");
int canMapHostMemory;
getCudaAttribute<int>(&canMapHostMemory, CU_DEVICE_ATTRIBUTE_CAN_MAP_HOST_MEMORY, dev);
printf(" Support host page-locked memory mapping: %s\n", canMapHostMemory ? "Yes" : "No");
int concurrentKernels;
getCudaAttribute<int>(&concurrentKernels, CU_DEVICE_ATTRIBUTE_CONCURRENT_KERNELS, dev);
printf(" Concurrent kernel execution: %s\n", concurrentKernels ? "Yes" : "No");
int surfaceAlignment;
getCudaAttribute<int>(&surfaceAlignment, CU_DEVICE_ATTRIBUTE_SURFACE_ALIGNMENT, dev);
printf(" Alignment requirement for Surfaces: %s\n", surfaceAlignment ? "Yes" : "No");
int eccEnabled;
getCudaAttribute<int>(&eccEnabled, CU_DEVICE_ATTRIBUTE_ECC_ENABLED, dev);
printf(" Device has ECC support: %s\n", eccEnabled ? "Enabled" : "Disabled");
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
int tccDriver;
getCudaAttribute<int>(&tccDriver, CU_DEVICE_ATTRIBUTE_TCC_DRIVER, dev);
printf(" CUDA Device Driver Mode (TCC or WDDM): %s\n",
tccDriver ? "TCC (Tesla Compute Cluster Driver)" : "WDDM (Windows Display Driver Model)");
#endif
int unifiedAddressing;
getCudaAttribute<int>(&unifiedAddressing, CU_DEVICE_ATTRIBUTE_UNIFIED_ADDRESSING, dev);
printf(" Device supports Unified Addressing (UVA): %s\n", unifiedAddressing ? "Yes" : "No");
int managedMemory;
getCudaAttribute<int>(&managedMemory, CU_DEVICE_ATTRIBUTE_MANAGED_MEMORY, dev);
printf(" Device supports Managed Memory: %s\n", managedMemory ? "Yes" : "No");
int computePreemption;
getCudaAttribute<int>(&computePreemption, CU_DEVICE_ATTRIBUTE_COMPUTE_PREEMPTION_SUPPORTED, dev);
printf(" Device supports Compute Preemption: %s\n", computePreemption ? "Yes" : "No");
int cooperativeLaunch;
getCudaAttribute<int>(&cooperativeLaunch, CU_DEVICE_ATTRIBUTE_COOPERATIVE_LAUNCH, dev);
printf(" Supports Cooperative Kernel Launch: %s\n", cooperativeLaunch ? "Yes" : "No");
int cooperativeMultiDevLaunch;
getCudaAttribute<int>(&cooperativeMultiDevLaunch, CU_DEVICE_ATTRIBUTE_COOPERATIVE_MULTI_DEVICE_LAUNCH, dev);
printf(" Supports MultiDevice Co-op Kernel Launch: %s\n", cooperativeMultiDevLaunch ? "Yes" : "No");
int pciDomainID, pciBusID, pciDeviceID;
getCudaAttribute<int>(&pciDomainID, CU_DEVICE_ATTRIBUTE_PCI_DOMAIN_ID, dev);
getCudaAttribute<int>(&pciBusID, CU_DEVICE_ATTRIBUTE_PCI_BUS_ID, dev);
getCudaAttribute<int>(&pciDeviceID, CU_DEVICE_ATTRIBUTE_PCI_DEVICE_ID, dev);
printf(" Device PCI Domain ID / Bus ID / location ID: %d / %d / %d\n", pciDomainID, pciBusID, pciDeviceID);
const char *sComputeMode[] = {"Default (multiple host threads can use ::cudaSetDevice() with device "
"simultaneously)",
"Exclusive (only one host thread in one process is able to use "
"::cudaSetDevice() with this device)",
"Prohibited (no host thread can use ::cudaSetDevice() with this "
"device)",
"Exclusive Process (many threads in one process is able to use "
"::cudaSetDevice() with this device)",
"Unknown",
NULL};
int computeMode;
getCudaAttribute<int>(&computeMode, CU_DEVICE_ATTRIBUTE_COMPUTE_MODE, dev);
printf(" Compute Mode:\n");
printf(" < %s >\n", sComputeMode[computeMode]);
}
}
printf("Result = PASS\n");
// If there are 2 or more GPUs, query to determine whether RDMA is supported
if (deviceCount >= 2) {
int gpuid[64]; // we want to find the first two GPUs that can support P2P
int gpu_p2p_count = 0;
int tccDriver = 0;
exit(EXIT_SUCCESS);
for (int i = 0; i < deviceCount; i++) {
checkCudaErrors(cuDeviceGetAttribute(&major, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR, i));
checkCudaErrors(cuDeviceGetAttribute(&minor, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR, i));
getCudaAttribute<int>(&tccDriver, CU_DEVICE_ATTRIBUTE_TCC_DRIVER, i);
// Only boards based on Fermi or later can support P2P
if ((major >= 2)
#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
// on Windows (64-bit), the Tesla Compute Cluster driver for windows
// must be enabled to support this
&& tccDriver
#endif
) {
// This is an array of P2P capable GPUs
gpuid[gpu_p2p_count++] = i;
}
}
// Show all the combinations of support P2P GPUs
int can_access_peer;
char deviceName0[256], deviceName1[256];
if (gpu_p2p_count >= 2) {
for (int i = 0; i < gpu_p2p_count; i++) {
for (int j = 0; j < gpu_p2p_count; j++) {
if (gpuid[i] == gpuid[j]) {
continue;
}
checkCudaErrors(cuDeviceCanAccessPeer(&can_access_peer, gpuid[i], gpuid[j]));
checkCudaErrors(cuDeviceGetName(deviceName0, 256, gpuid[i]));
checkCudaErrors(cuDeviceGetName(deviceName1, 256, gpuid[j]));
printf("> Peer-to-Peer (P2P) access from %s (GPU%d) -> %s (GPU%d) : "
"%s\n",
deviceName0,
gpuid[i],
deviceName1,
gpuid[j],
can_access_peer ? "Yes" : "No");
}
}
}
}
printf("Result = PASS\n");
exit(EXIT_SUCCESS);
}

1
Samples/1_Utilities/topologyQuery/README.md
View File

@@ -30,4 +30,3 @@ cudaGetDeviceCount, cudaDeviceGetAttribute
Download and install the [CUDA Toolkit 12.5](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
## References (for more details)

68
Samples/1_Utilities/topologyQuery/topologyQuery.cu
View File

@@ -35,48 +35,44 @@
// includes, project
#include <helper_cuda.h>
#include <helper_functions.h> // helper for shared that are common to CUDA Samples
#include <helper_functions.h> // helper for shared that are common to CUDA Samples
int main(int argc, char **argv) {
int deviceCount = 0;
checkCudaErrors(cudaGetDeviceCount(&deviceCount));
int main(int argc, char **argv)
{
int deviceCount = 0;
checkCudaErrors(cudaGetDeviceCount(&deviceCount));
// Enumerates Device <-> Device links
for (int device1 = 0; device1 < deviceCount; device1++) {
for (int device2 = 0; device2 < deviceCount; device2++) {
if (device1 == device2) continue;
// Enumerates Device <-> Device links
for (int device1 = 0; device1 < deviceCount; device1++) {
for (int device2 = 0; device2 < deviceCount; device2++) {
if (device1 == device2)
continue;
int perfRank = 0;
int atomicSupported = 0;
int accessSupported = 0;
int perfRank = 0;
int atomicSupported = 0;
int accessSupported = 0;
checkCudaErrors(cudaDeviceGetP2PAttribute(
&accessSupported, cudaDevP2PAttrAccessSupported, device1, device2));
checkCudaErrors(cudaDeviceGetP2PAttribute(
&perfRank, cudaDevP2PAttrPerformanceRank, device1, device2));
checkCudaErrors(cudaDeviceGetP2PAttribute(
&atomicSupported, cudaDevP2PAttrNativeAtomicSupported, device1,
device2));
checkCudaErrors(
cudaDeviceGetP2PAttribute(&accessSupported, cudaDevP2PAttrAccessSupported, device1, device2));
checkCudaErrors(cudaDeviceGetP2PAttribute(&perfRank, cudaDevP2PAttrPerformanceRank, device1, device2));
checkCudaErrors(
cudaDeviceGetP2PAttribute(&atomicSupported, cudaDevP2PAttrNativeAtomicSupported, device1, device2));
if (accessSupported) {
std::cout << "GPU" << device1 << " <-> GPU" << device2 << ":"
<< std::endl;
std::cout << " * Atomic Supported: "
<< (atomicSupported ? "yes" : "no") << std::endl;
std::cout << " * Perf Rank: " << perfRank << std::endl;
}
if (accessSupported) {
std::cout << "GPU" << device1 << " <-> GPU" << device2 << ":" << std::endl;
std::cout << " * Atomic Supported: " << (atomicSupported ? "yes" : "no") << std::endl;
std::cout << " * Perf Rank: " << perfRank << std::endl;
}
}
}
}
// Enumerates Device <-> Host links
for (int device = 0; device < deviceCount; device++) {
int atomicSupported = 0;
checkCudaErrors(cudaDeviceGetAttribute(
&atomicSupported, cudaDevAttrHostNativeAtomicSupported, device));
std::cout << "GPU" << device << " <-> CPU:" << std::endl;
std::cout << " * Atomic Supported: " << (atomicSupported ? "yes" : "no")
<< std::endl;
}
// Enumerates Device <-> Host links
for (int device = 0; device < deviceCount; device++) {
int atomicSupported = 0;
checkCudaErrors(cudaDeviceGetAttribute(&atomicSupported, cudaDevAttrHostNativeAtomicSupported, device));
std::cout << "GPU" << device << " <-> CPU:" << std::endl;
std::cout << " * Atomic Supported: " << (atomicSupported ? "yes" : "no") << std::endl;
}
return 0;
return 0;
}