07-30, 13:30–14:00 (UTC), Green Track
KernelAbstractions.jl is a library to write portable kernels for CPU and GPU computing.
It presents a unified programming model ala OpenCL in Julia so that users can write a
single kernel that targets multiple accelerators an multi-threaded CPUs.
Writing performance portable code that targets multiple hardware platforms is challenging.
KernelAbstractions.jl allows users to abstract away some of the hardware differences and
write one unified code that targets multiple hardware platforms. Despite being able to target
multi-threaded CPUs it's design choices are biased towards GPUs and can sometimes be sub-optimal
for CPU execution. In this talk I will introduce KernelAbstractions.jl, and explain it's design choices.
