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

We present a new algorithm to quickly generate high-performance GPUimplementations of complex imaging and vision pipelines, directly fromhigh-level Halide algorithm code. It is fully automatic, requiring no scheduletemplates or hand-optimized kernels. We address the scalability challenge ofextending search-based automatic scheduling to map large real-world programs tothe deep hierarchies of memory and parallelism on GPU architectures inreasonable compile time. We achieve this using (1) a two-phase search algorithmthat first 'freezes' decisions for the lowest cost sections of a program,allowing relatively more time to be spent on the important stages, (2) ahierarchical sampling strategy that groups schedules based on their structuralsimilarity, then samples representatives to be evaluated, allowing us toexplore a large space with few samples, and (3) memoization of repeated partialschedules, amortizing their cost over all their occurrences. We guide theprocess with an efficient cost model combining machine learning, programanalysis, and GPU architecture knowledge. We evaluate our method's performanceon a diverse suite of real-world imaging and vision pipelines. Our scalabilityoptimizations lead to average compile time speedups of 49x (up to 530x). Wefind schedules that are on average 1.7x faster than existing automaticsolutions (up to 5x), and competitive with what the best human experts wereable to achieve in an active effort to beat our automatic results.

Efficient Automatic Scheduling of Imaging and Vision Pipelines for the GPU