I previously used NV Visual Profiler with CUDA 5.0 to view the GPU usage of the OptiX application I’m working on. Now that I’ve upgraded to CUDA 6.0 and am using the version of NVVP that comes with it, I can no longer see entries for my GPUs in the timeline. Instead, I only see entries for the Runtime API, Driver API, and Profiling Overhead. The timeline also stops early, usually after a few seconds or milliseconds.
In the Analysis tab, I see a message saying “No GPU devices in session, GPU usage analysis cannot be performed.”
I have the same problem with many (though not all) of the CUDA Samples.
Any ideas how to fix this? I’m running Windows 7 and have tried 32-bit and 64-bit versions of my program.
maybe your kernel launches are failing. What happens if you run your app with cuda-memcheck?
Also, maybe your CUDA setup is now broken. What happens if you run
nvidia-smi -a
?
What is the output of running the deviceQuery sample app?
My program does generate an output image, which leads me to believe that the kernel launch is not failing. nvidia-smi -a shows the output I would expect for both of the machines I’ve tried this on (one with a Quadro K4000, one with two Tesla K40s).
Since you asked, here’s the output from deviceQuery:
CUDA Device Query (Runtime API) version (CUDART static linking)
Detected 1 CUDA Capable device(s)
Device 0: "Quadro K4000"
CUDA Driver Version / Runtime Version 6.0 / 6.0
CUDA Capability Major/Minor version number: 3.0
Total amount of global memory: 3072 MBytes (3221225472 bytes)
( 4) Multiprocessors, (192) CUDA Cores/MP: 768 CUDA Cores
GPU Clock rate: 811 MHz (0.81 GHz)
Memory Clock rate: 2808 Mhz
Memory Bus Width: 192-bit
L2 Cache Size: 393216 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(65536), 2D=(65536, 65536), 3D=(4096, 4096, 4096)
Maximum Layered 1D Texture Size, (num) layers 1D=(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 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 = 6.0, CUDA Runtime Version = 6.0, NumDevs = 1, Device0 = Quadro
K4000
Result = PASS
I have find the solution for this problem,you just need to add the function cudaProfilerStop() behind your program.Don’t forget the head file #include <cuda_profiler_api.h>
I have found why! Before using visual Profiler to analyse project, you should add two functions before codes which execute in kernel function and after: cudaProfilerStart(); and cudaProfilerStop().