Vikunja is a performance portable algorithm library that defines functions operating on ranges of elements for a variety of purposes . It supports the execution on multi-core CPUs and various GPUs. Vikunja uses alpaka to implement platform-independent primitives such as reduce or transform.
Currently, only the device-reduce is provided as a ready-to-use function.
The block-reduce exists as a kernel, but it might be useful to provide this as a convenient, ready-to-use function like CUB does.
Currently, only the device-reduce is provided as a ready-to-use function. The block-reduce exists as a kernel, but it might be useful to provide this as a convenient, ready-to-use function like CUB does.