2017
DOI: 10.1145/3162017
|View full text |Cite
|
Sign up to set email alerts
|

Optimizing Affine Control With Semantic Factorizations

Abstract: Hardware accelerators generated by polyhedral synthesis make an extensive use of affine expressions (affine functions and convex polyhedra) in control and steering logic. Since the control is pipelined, these affine objects must be evaluated at the same time for different values, which forbids aggressive reuse of operators. In this report, we propose an algorithm to factorize a collection of affine expressions without preventing pipelining. Our key contributions are (i) to use semantic factorizations exploitin… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2018
2018
2021
2021

Publication Types

Select...
2

Relationship

2
0

Authors

Journals

citations
Cited by 2 publications
(3 citation statements)
references
References 38 publications
0
3
0
Order By: Relevance
“…With the PPN partitioning strategy, it is possible to factor the channels [43] and the processes which share a lot of common control. We also proposed a back-end algorithm to compact affine control [6]. Section 6 will outline our back-end.…”
Section: Compilation Methodologymentioning
confidence: 99%
See 1 more Smart Citation
“…With the PPN partitioning strategy, it is possible to factor the channels [43] and the processes which share a lot of common control. We also proposed a back-end algorithm to compact affine control [6]. Section 6 will outline our back-end.…”
Section: Compilation Methodologymentioning
confidence: 99%
“…The Inmux is responsible of fetching data from the Buffers and Demuxes to store the data into output Buffers. All these units make an extensive use of piece-wise affine functions, which are efficiently synthesized as DAGs, using the technique described in [6]. These units communicate over a long pipelined bus with channels.…”
Section: Back-endmentioning
confidence: 99%
“…The bigger is a parallel unit, the less it can be duplicated, thereby limiting the overall performance. Particularly, tricky program optimizations are likely to spoil the performances if the circuit is not post-optimized carefully [5]. An important consequence is that the the roofline model is not longer valid in HLS [8].…”
Section: Introductionmentioning
confidence: 99%