npm install @rapidsai/cuda)A node native addon that provides bindings to the CUDA driver and runtime APIs.
These bindings allow calling the CUDA device management, memory, stream, event, ipc, and
DX/GL interop APIs from JS. These APIs allow node (v8 or chakracore) applications to read,
write, and share memory via zero-copy CUDA IPC with external processes that also use the
CUDA, OpenGL, and RAPIDS libraries.
cudaChooseDevice, cudaGetDeviceCount, cuDeviceGet, cudaDeviceGetPCIBusId, cudaDeviceGetByPCIBusId, cudaGetDevice, cudaGetDeviceFlags, cudaGetDeviceProperties, cudaSetDevice, cudaSetDeviceFlags, cudaDeviceReset, cudaDeviceSynchronize, cudaDeviceCanAccessPeer, cudaDeviceEnablePeerAccess, cudaDeviceDisablePeerAccess
CUDADevice: A class to wrap and manage a CUDA device.CUDABuffer: A class to wrap and manage device memory allocations (similar to ArrayBuffer).CUDAArray: A class to wrap operations like read/write/share on a CUDABuffer (similar to TypedArray).cuPointerGetAttribute, cudaMalloc, cudaFree, cudaMallocHost, cudaFreeHost, cudaHostRegister, cudaHostUnregister, cudaMemcpy, cudaMemset, cudaMemcpyAsync, cudaMemsetAsync, cudaMemGetInfo
cudaIpcGetMemHandle, cudaIpcOpenMemHandle, cudaIpcCloseMemHandle
cudaStreamCreate, cudaStreamDestroy, cudaStreamSynchronize
cuGraphicsGLRegisterBuffer, cuGraphicsGLRegisterImage, cuGraphicsUnregisterResource, cuGraphicsMapResources, cuGraphicsUnapResources, cuGraphicsResourceGetMappedPointer
A typed array of 64-bit floating point numbers values in CUDA memory.
A typed array of twos-complement 16-bit signed integers in CUDA memory.
A typed array of twos-complement 32-bit signed integers in CUDA memory.
A typed array of 64-bit signed integers in CUDA memory.
A typed array of twos-complement 8-bit signed integers in CUDA memory.
A container for managing the lifetime of a DeviceMemory allocation exported for reading and/or writing by other processes with access to the allocation's associated Device.
A base class for typed arrays of values in owned or managed by CUDA.
A typed array of 16-bit unsigned integers in CUDA memory.
A typed array of 32-bit unsigned integers in CUDA memory.
A typed array of 64-bit unsigned integers in CUDA memory.
A typed array of 8-bit unsigned integers in CUDA memory.
A typed array of 8-bit unsigned integers clamped to 0-255 in CUDA memory.
A lazily-evaluated list of available CUDA devices.
This list has a length property, and each available active Device can be accessed by device
ordinal (via Array-style subscript-accesses).
This list implements the Iterablefor..of
loop, or with the [...] iterable expansion syntax.
Clamp begin and end ranges similar to Array.prototype.slice.
A function to override the default device memory allocation behavior.
The supplied function will be called to create the underlying Memory instances
when constructing one of the CUDAMemoryView in JavaScript.
A typed array of 32-bit floating point numbers in CUDA memory.