Giters

alpaka-group / vikunja

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.

Home Page:https://vikunja.readthedocs.io/en/latest/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Verify manuall loop unrolling in BlockThreadReduceKernel and overwork it

SimeonEhrig opened this issue · comments

commented

There is comment BlockThreadReduceKernel, that the manual unrolling is not tested.

vikunja/include/vikunja/reduce/detail/BlockThreadReduceKernel.hpp

Line 114 in b3e3dc3

// Manual unrolling. I dont know if this is really necessary, but

This should be done and @bernhardmgruber mentioned in the PR #40 that some parts could be illegal and provide some improvements, also for the readability.

Related to the topic: #40 (comment)