Enhancing High-Level Synthesis Using a Meta-Programming Approach

Vandebon, Jessica; Coutinho, Jose de Figueiredo; Luk, Wayne; Nurvitadhi, Eriko

doi:10.1109/tc.2021.3096429

Cited by 7 publications

(4 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2(b) illustrates two specific envisioned roles: (1) the application developer whose expertise and primary concern is algorithmic behaviour; and (2) the design-flow developer who builds designflows that automate mapping and optimisation. This is possible through our meta-programming methodology [21], which allows programs to be seen as data, and therefore programmatically analysed and manipulated. This means that all currently manual designflow tasks, from analysis to code transforms and DSE, can be potentially codified and coordinated into an end-to-end design-flow that operates on a high-level unoptimised software description to derive an optimised design with little intervention from an application developer.…”

Section: Design-flow Patterns 21 Overviewmentioning

confidence: 99%

“…We implement the design-flow patterns in our catalogue using the meta-programming tool Artisan [21]. Artisan provides a unified environment for source-code analysis, instrumentation, and execution, with support for an array of key features that facilitate modular pattern implementation as well as coordination into end-to-end design-flows.…”

Section: Codifying Design-flow Patterns 31 the Artisan Meta-programmi...mentioning

confidence: 99%

“…The Artisan meta-programming framework has previously been used to programmatically optimise FPGA designs starting from sequential C++ source-code [21], with efficient, customisable, and reusable optimisation strategies automated using meta-programs in order to decouple behavioural application descriptions from optimisation descriptions. In this paper, we employ this framework to support design-flow patterns, while providing a higher-level of abstraction above meta-program implementations, broadening the scope for reasoning about modular components and their coordination into design-flows for different targets, not tied to a particular implementation.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Meta-Programming Design-Flow Patterns for Automating Reusable Optimisations

Vandebon

Coutinho

Luk

2022

International Symposium on Highly-Efficient Accelerators and Reconfigurable Technologies

Self Cite

View full text Add to dashboard Cite

Continuing advances in heterogeneous and parallel computing enable massive performance gains in domains such as AI and HPC. Such gains often involve using hardware accelerators, such as FP-GAs and GPUs, to speed up specific workloads. However, to make effective use of emerging heterogeneous architectures, optimisation is typically done manually by highly-skilled developers with in-depth understanding of the target hardware. The process is tedious, error-prone, and must be repeated for each new application. This paper introduces Design-Flow Patterns, which capture modular, recurring application-agnostic elements involved in mapping and optimising application descriptions onto efficient CPU and GPU targets. Our approach is the first to codify and programmatically coordinate these elements into fully automated, customisable, and reusable end-to-end design-flows. We implement key design-flow patterns using the meta-programming tool Artisan, and evaluate automated design-flows applied to three sequential C++ applications. Compared to single-threaded implementations, our approach generates multi-threaded OpenMP CPU designs achieving up to 18 times speedup on a CPU platform with 32-threads, as well as HIP GPU designs achieving up to 1184 times speedup on an NVIDIA GeForce RTX 2080 Ti GPU. CCS CONCEPTS• Computer systems organization → Heterogeneous (hybrid) systems; Multicore architectures; • Software and its engineering → Compilers.

show abstract

Section: Design-flow Patterns 21 Overviewmentioning

confidence: 99%

Section: Codifying Design-flow Patterns 31 the Artisan Meta-programmi...mentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Meta-Programming Design-Flow Patterns for Automating Reusable Optimisations

Vandebon

Coutinho

Luk

2022

International Symposium on Highly-Efficient Accelerators and Reconfigurable Technologies

Self Cite

View full text Add to dashboard Cite

show abstract

“…This mining method has brought great help to people and has become one of the aspects that many people are willing to explore and develop. Among them, Python is a popular program language in this field, because it has a lot of content, powerful technology and powerful computing power [2].…”

Section: Introductionmentioning

confidence: 99%

Clustering Mining Algorithm of Internet of Things Database Based on Python Language

Wan,

Liu

2023

cai

View full text Add to dashboard Cite

In order to solve the problems of reading delay in data mining of the Internet of Things database, a clustering mining algorithm of the Internet of Things database based on Python language is proposed. We designed an improved crawler algorithm based on the open-source structure of scratch through Python language, judge the similarity of recruitment data topics in the Internet of Things database through Bayesian classifier, and crawl the recruitment data in the Internet of Things database: the number of keywords in the text space, the degree of keyword extraction, and the number of keyword data in the text space. The time series model is used to eliminate the delay of text features. On this basis, the semi-supervised learning and semi-cluster analysis method is used to construct the corresponding classifier, complete the adaptive classification process of the text data stream and realize the clustering mining of the Internet of Things database based on Python language. The experimental results show that this method has a low reading delay, and can mine the attention, number of posts and click time frequency of the Internet of Things database from which the recruitment data are obtained.

show abstract