2019 International Conference on Field-Programmable Technology (ICFPT) 2019
DOI: 10.1109/icfpt47387.2019.00013
|View full text |Cite
|
Sign up to set email alerts
|

SkyCastle: A Resource-Aware Multi-Loop Scheduler for High-Level Synthesis

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2

Citation Types

0
2
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
2
1
1

Relationship

1
3

Authors

Journals

citations
Cited by 4 publications
(2 citation statements)
references
References 20 publications
0
2
0
Order By: Relevance
“…During such a process, it is important to be able to quickly assess the accelerator, not only its performance, but also with regard to hardware resource consumption and operating frequency. In their research on high-level synthesis scheduling algorithms [52,53], Oppermann et al used TaPaSCo for exactly this purpose: Harnessing the easy-to-use out-of-context synthesis feature that is part of TaPaSCo's high-level synthesis and import flow (cf. Section 3.4), it was possible easily to evaluate a large number of examples and scheduler configurations with regard to effects of the scheduling algorithm on resource requirements and the critical path delay in hardware.…”
Section: Integration Of High-level Synthesismentioning
confidence: 99%
“…During such a process, it is important to be able to quickly assess the accelerator, not only its performance, but also with regard to hardware resource consumption and operating frequency. In their research on high-level synthesis scheduling algorithms [52,53], Oppermann et al used TaPaSCo for exactly this purpose: Harnessing the easy-to-use out-of-context synthesis feature that is part of TaPaSCo's high-level synthesis and import flow (cf. Section 3.4), it was possible easily to evaluate a large number of examples and scheduler configurations with regard to effects of the scheduling algorithm on resource requirements and the critical path delay in hardware.…”
Section: Integration Of High-level Synthesismentioning
confidence: 99%
“…Evaluated using the Rodinia benchmark, their model present an average percentage error of 9.5% and 12.6% for performance and power, respectively, and can explore design of spaces of 128 points within 3 and 6 minutes. Oppermann et al (2019) present SkyCastle, a resource-aware multi-loop HLS scheduler. SkyCastle tackles loop pipelining using an Integer Linear Programming method to pipeline multiple loop nests.…”
Section: High-level Synthesis Optimisation Frameworkmentioning
confidence: 99%