English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

There is a large body of legacy scientific code written in languages likeFortran that is not optimised to get the best performance out of heterogeneousacceleration devices like GPUs and FPGAs, and manually porting such code intoparallel languages frameworks like OpenCL requires considerable effort. We areworking towards developing a turn-key, self-optimising compiler foraccelerating scientific applications, that can automatically transform legacycode into a solution for heterogeneous targets. In this paper we focus on FPGAsas the acceleration devices, and carry out our discussion in the context of theOpenCL programming framework. We show a route to automatic creation of kernelswhich are optimised for execution in a "streaming" fashion, which gives optimalperformance on FPGAs. We use a 2D shallow-water model as an illustration;specifically we show how the use of \emph{channels} to communicate directlybetween peer kernels and the use of on-chip memory to create stencil bufferscan lead to significant performance improvements. Our results show better FPGAperformance against a baseline CPU implementation, and better energy-efficiencyagainst both CPU and GPU implementations.

Towards Automatic Transformation of Legacy Scientific Code into OpenCL for Optimal Performance on FPGAs