OrangeOwlSolutions / Cuda Utilities
Utilities for CUDA programming
Stars: ✭ 30
Labels
Projects that are alternatives of or similar to Cuda Utilities
Neuralsuperresolution
Real-time video quality improvement for applications such as video-chat using Perceptual Losses
Stars: ✭ 18 (-40%)
Mutual labels: cuda
Graphvite
GraphVite: A General and High-performance Graph Embedding System
Stars: ✭ 865 (+2783.33%)
Mutual labels: cuda
Uammd
A CUDA project for Molecular Dynamics, Brownian Dynamics, Hydrodynamics... intended to simulate a very generic system constructing a simulation with modules.
Stars: ✭ 11 (-63.33%)
Mutual labels: cuda
Sepconv Slomo
an implementation of Video Frame Interpolation via Adaptive Separable Convolution using PyTorch
Stars: ✭ 918 (+2960%)
Mutual labels: cuda
Imagenet Classifier Tensorflow
Image recognition and classification using Convolutional Neural Networks with TensorFlow
Stars: ✭ 13 (-56.67%)
Mutual labels: cuda
Wheels
Performance-optimized wheels for TensorFlow (SSE, AVX, FMA, XLA, MPI)
Stars: ✭ 891 (+2870%)
Mutual labels: cuda
Theano Roi Align
An implementation of the RoiAlign operation for Theano
Stars: ✭ 11 (-63.33%)
Mutual labels: cuda
CUDA Utilities
Utilities for CUDA programming
Wrap all the CUDA API calls within gpuErrchk
. For example:
gpuErrchk(cudaMemcpy(d_x, x, N * sizeof(float), cudaMemcpyHostToDevice));
Use gpuErrchk
to check for errors in your kernels. For example:
#ifdef DEBUG
gpuErrchk(cudaPeekAtLastError());
gpuErrchk(cudaDeviceSynchronize());
#endif
If DEBUG
is defined, then the CUDA error check of the kernels will be executed.
Use iDivUp
to find the number of blocks to be launched by a CUDA kernel. For example:
BLOCKSIZE = 256;
N = 10000; // Number of elements to be processed
kernel<<<iDivUp(N, BLOCKSIZE), BLOCKSIZE>>>(...);
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].