Futhark is written in Haskell.
This includes the ability to support nested parallelizations (parallel operations inside other parallel operations).
The researchers claims are not unfounded.Preliminary benchmarkspitting Futhark GPU programs against other GPU libraries showed promising results.
For a maximum segment sum test, Futhark was nearly twice as fast as a competing open source project.
In other cases, it matched or was only slightly slower than the competition.
Futharks developers attributed this to the newness of the language and its unrefined approach to allocating and reusing memory.
Python users who want a speed boost can generate code with Futhark wrapped automatically in a Python module.
The GPU code runs viaPyOpenCL, which is a library for interfacing with GPUs.
There areexamplesof this, such as a Futhark-drivenGPU implementation of Conways Game of Life.
source: www.techworm.net
