Kernel Operations on the GPU, with Autodiff, without Memory Overflows.
Benjamin CharlierJean FeydyJoan Alexis GlaunèsFrançois-David CollinGhislain DurifPublished in: CoRR (2020)
Keyphrases
- memory requirements
- kernel function
- storage media
- kernel methods
- memory space
- real time
- neural network
- graphics processors
- parallel processing
- general purpose
- feature space
- main memory
- gaussian processes
- machine learning
- kernel matrix
- memory usage
- random access
- memory size
- multi threaded
- mutual subspace method
- memory bandwidth
- intel xeon
- processing elements
- graph kernels
- kernel machines
- parallel computation
- computational power
- parallel implementation
- parallel algorithm
- support vector
- feature selection