gpu

问题 For loops seem to be extremely slow, so I was wondering if the nested loops in the code shown next could be vectorized using bsxfun and maybe GPU could be introduced too. Code %// Paramaters i = 1; j = 3; n1 = 1500; n2 = 1500; %// Pre-allocate for output LInc(n1+n2,n1+n2)=0; %// Nested Loops - I for x = 1:n1 for y = 1:n1 num = ((n2 ^ 2) * (L1(i, i) + L2(j, j) + 1)) - (n2 * n * (L1(x,i) + L1(y,i))); LInc(x, y) = L1(x, y) + (num/denom); LInc(y, x) = LInc(x, y); end end %// Nested Loops - II for

问题 i'm training some Music Data on a LSTM-RNN in Tensorflow and encountered some Problem with GPU-Memory-Allocation which i don't understand: I encounter an OOM when there actually seems to be just about enough VRAM still available. Some background: I'm working on Ubuntu Gnome 16.04, using a GTX1060 6GB, Intel Xeon E3-1231V3 and 8GB RAM. So now first the part of the error-message which i can understand, in the and i will add the whole error message in the end again for anyone who might ask for

问题 Im trying to increase the timeout on the GPU from its default setting of 2 seconds to something a little longer. I found the following link but it appears its slightly different in windows 7 as i cant see anything mentioned in the webpage. Has anyone done this before? If so could you fill in the gaps please. Thanks @RoBik so as follows if i want 6 days (bit excessive i know but just for example)? Thanks again for your help, +1. EDIT This is the error im currently getting. An error has occured

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 2 days ago . I am a beginner in cuda, I already executed the command of enabling L1 Cache through command prompt in windows. nvcc locates in the path: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\bin on my device. I write the command as follows. C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\bin>nvcc

问题 lets consider the following array: tab = [80,12,14,5,70,9,26,30,8,12,16,15] I want to compute the sum of all possible sequences of size 4 using cuda : for example : S1=80+12+14+5=111 S2=12+14+5+70 =101 S3=14+5+70+9 =98 .... You have an efficient idea to parallise this task using Cuda. the previous table is just an example in my case i will use huge one. 回答1: We can do this in a single operation ( thrust::transform ) using thrust. In CUDA, this can be considered to be a fairly simple 1-D

问题 I have CUDA 9.0 and CUDNN 7.1 installed on Ubuntu 18.04(Linux mint 19). Tensorflow-gpu works fine on GPU(GTX 1080ti). Now i am trying to build dlib with CUDA support: sudo python3 setup.py install --yes USE_AVX_INSTRUCTIONS --yes DLIB_USE_CUDA --clean Got the error: user@user-pc:~/Downloads/dlib$ sudo python3 setup.py install --yes USE_AVX_INSTRUCTIONS --yes DLIB_USE_CUDA --clean running install running bdist_egg running egg_info writing dlib.egg-info/PKG-INFO writing dependency_links to dlib

问题 This is my from my MATLAB script. function [ Im ] = findBorders( I ) Im = false(size(I)); I = padarray(I, [1, 1], 1); [h w] = size(Im); bkgFound = false; for row = 1 : h for col = 1 : w if I(row + 1, col + 1) bkgFound = false; for i = 0:2 for j = 0:2 if ~I(row + i, col + j) Im(row, col) = 1; bkgFound = true; break; end; end; if bkgFound break; end; end; end; end; end; end So, I need to convert it to parfor loop, to run into GPU. I need help. I read some articles, but have no idea about how to

问题 My CUDA kernel function is not returning the intended result (a sum of all elements in vector b) but is instead returning a single value from vector b. I tried memcheck and racecheck, but nothing came up: [breecej@compute-0-32 newsum]$ cuda-memcheck mystock ========= CUDA-MEMCHECK ========= ERROR SUMMARY: 0 errors [breecej@compute-0-32 newsum]$ cuda-memcheck --tool racecheck mystock ========= CUDA-MEMCHECK ========= RACECHECK SUMMARY: 0 hazards displayed (0 errors, 0 warnings) [breecej

问题 I have got a cluster with gpu nodes (nvidia) and deployed DC/OS 1.8. I'd like to enable to schedule jobs (batch and spark) on gpu nodes using gpu isolation. DC/OS is based on mesos 1.0.1 that supports gpu isolation. 回答1: Unfortunately, DC/OS doesn't officially support GPUs in 1.8 ( experimental support for GPUs will be coming in the next release as mentioned here: https://github.com/dcos/dcos/pull/766 ). In this next release, only Marathon will officially be able to launch GPU services

问题 I have got a cluster with gpu nodes (nvidia) and deployed DC/OS 1.8. I'd like to enable to schedule jobs (batch and spark) on gpu nodes using gpu isolation. DC/OS is based on mesos 1.0.1 that supports gpu isolation. 回答1: Unfortunately, DC/OS doesn't officially support GPUs in 1.8 ( experimental support for GPUs will be coming in the next release as mentioned here: https://github.com/dcos/dcos/pull/766 ). In this next release, only Marathon will officially be able to launch GPU services