nvidia

问题 In the nvidia driver package, there is libcuda.so. Is the cuda driver the same as nvidia driver? And what's the relation between cuda toolkit and libcuda.so? 回答1: From cuda document http://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#driver-api http://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#versioning-and-compatibility It seems cuda driver is libcuda.so which is included in nvidia driver and used by cuda runtime api Nvidia driver includes driver kernel module and

问题 This is on windows 10 computer with no monitor attached to the Nvidia card. I've included output from nvida-smi showing > 5.04G was available. Here is the tensorflow code asking it to allocate just slightly more than I had seen previously: (I want this to be as close as possible to memory fraction=1.0) config = tf.ConfigProto() #config.gpu_options.allow_growth=True config.gpu_options.per_process_gpu_memory_fraction=0.84 config.log_device_placement=True sess = tf.Session(config=config) Just

问题 I starting to implement custum video decoder that utilize cuda HW decoder to generate YUV frame for next to encode it. How can I fill "CUVIDPICPARAMS" struc ??? Is it possible? My algorithm are: For get video stream packet I'm use ffmpeg-dev libs avcodec, avformat... My steps: 1) Open input file: avformat_open_input(&ff_formatContext,in_filename,nullptr,nullptr); 2) Get video stream property's: avformat_find_stream_info(ff_formatContext,nullptr); 3) Get video stream: ff_video_stream=ff

问题 I am working on a stochastic process and I wanted to generate different series if random numbers in CUDA kernel each time I run the program. This similar to what we does in C++ by declaring seed = time(null) followed by srand(seed) and rand( ) I can pass seeds from host to device via the kernel but the problem in doing this is I would have to pass an entire array of seeds into the kernel for each thread to have a different random seed each time. Is there a way I could generate random seed /

问题 Closed. This question is off-topic. It is not currently accepting answers. Want to improve this question? Update the question so it's on-topic for Stack Overflow. Closed 4 years ago . In the documentation for CUDA 6.5 has written: http://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#ixzz3PIXMTktb 5.2.3. Multiprocessor Level ... 8L for devices of compute capability 3.x since a multiprocessor issues a pair of instructions per warp over one clock cycle for four warps at a time, as

问题 I would like to use NVIDIA TensorRT to run my Tensorflow models. Currenly, TensorRT supports Caffe prototxt network descriptor files. I was not able to find source code to convert Tensorflow models to Caffe models. Are there any workarounds? 回答1: TensorRT 3.0 supports import/conversion of TensorFlow graphs via it's UFF (universal framework format). Some layer implementations are missing and will require custom implementations via IPlugin interface. Previous versions didn't support native

问题 I have a 4 GPU machine on which I run Tensorflow (GPU) with Keras. Some of my classification problems take several hours to complete. nvidia-smi returns Volatile GPU-Util which never exceeds 25% on any of my 4 GPUs. How can I increase GPU Util% and speed up my training? 回答1: If your GPU util is below 80%, this is generally the sign of an input pipeline bottleneck. What this means is that the GPU sits idle much of the time, waiting for the CPU to prepare the data: What you want is the CPU to

问题 everybody, I have the following problem: I started a P2 instance with this AMI. I installed some tools like screen, torch, etc. Then I successfully run some experiments using GPU and I created an image of the instance, so that I can terminate it and run it again later. Later I started a new instance from the AMI I created before. Everything looked fine - screen, torch, my experiments were present on the system, but I couldn't run the same experiments as before: NVIDIA-SMI has failed because

问题 everybody, I have the following problem: I started a P2 instance with this AMI. I installed some tools like screen, torch, etc. Then I successfully run some experiments using GPU and I created an image of the instance, so that I can terminate it and run it again later. Later I started a new instance from the AMI I created before. Everything looked fine - screen, torch, my experiments were present on the system, but I couldn't run the same experiments as before: NVIDIA-SMI has failed because

问题 __global__ void helloCUDA(float f) { printf("Hello thread %d, f=%f\n", threadIdx.x, f); } int main() { helloCUDA<<<1, 5>>>(1.2345f); cudaDeviceSynchronize(); return 0; } Why is cudaDeviceSynchronize(); at many places for example here it is not required after kernel call? 回答1: A kernel launch is asynchronous . This means it returns control to the CPU thread immediately after starting up the GPU process, before the kernel has finished executing. So what is the next thing in the CPU thread here?