|
Mila
Deep Neural Network Library
|
Memory management layer for tensor data with support for CPU and GPU storage. More...
#include <cuda_runtime.h>#include <memory>#include <vector>#include <limits>#include <iostream>#include <stdexcept>#include <cstring>#include <type_traits>import Cuda.Error;import Compute.CudaMemoryResource;import Compute.CpuMemoryResource;import Compute.MemoryResourceTracker;import Compute.MemoryResource;Classes | |
| class | Mila::Dnn::TensorBuffer< T, MR, TrackMemory > |
| A buffer for storing tensor data with configurable memory management. More... | |
Namespaces | |
| namespace | Mila |
| namespace | Mila::Dnn |
| namespace | Mila::Dnn::detail |
Functions | |
| template<typename MR > | |
| constexpr size_t | Mila::Dnn::detail::get_alignment () |
| Determines the appropriate memory alignment based on the memory resource type. | |
Variables | |
| constexpr size_t | Mila::Dnn::detail::CPU_SIMD_ALIGN = 64 |
| AVX-512 alignment for CPU operations. | |
| constexpr size_t | Mila::Dnn::detail::CUDA_WARP_SIZE = 32 |
| CUDA warp size alignment (32 threads) | |
Memory management layer for tensor data with support for CPU and GPU storage.