Paper Info
Title
SPARK: A High-Lev l Synthesis Framework For Applying Parallelizing Compiler Transformations
Abstract
This paper presents a modular and extensible high-levelsynthesis research system, called SPARK, that takes a behavioral description in ANSI-C as input and produces synthesizable register-transfer level VHDL. SPARK uses parallelizing compiler technology developed previously to enhance instruction-level parallelism and re-instruments it forhigh-level synthesis by incorporating ideas of mutual exclusivity of operations, resource sharing and hardware costmodels. In this paper, we present the design flow throughthe SPARK system, a set of transformations that includespeculative code motions and dynamic transformations andshow how these transformations and other optimizing synthesis and compiler techniques are employed by a scheduling heuristic. Experiments are performed on two moderately complex industrial applications, namely, MPEG-1 andthe GIMP image processing tool.The results show that thevarious code transformations lead to up to 70% improvements in performance without any increase in the overallarea and critical path of the final synthesized design.
Year
DOI
Venue
2003
10.1109/ICVD.2003.1183177
VLSI Design
Keywords
Field
DocType
final synthesized design,optimizing synthesis,includespeculative code motion,design flow throughthe SPARK,compiler technology,thevarious code transformation,extensible high-levelsynthesis research system,forhigh-level synthesis,compiler technique,dynamic transformations andshow,High-Lev l,Parallelizing Compiler Transformations
Spark (mathematics),Computer science,Parallel computing,High-level synthesis,Real-time computing,Compiler,Design flow,Critical path method,Modular design,VHDL,Hardware description language
Conference
ISBN
Citations 
PageRank 
0-7695-1868-0
83
3.68
References 
Authors
11
4
Name
Order
Citations
PageRank
Sumit Gupta122113.28
Nikil Dutt24960421.49
Rajesh Gupta31024.98
Alex Nicolau485856.64