Multiloop Parallelisation Using Unrolling and Fission

Table 1

Some approaches to address mapping/scheduling.

References	Approach	Examples of applications	Comments
[1, 2]	Control flow based	GCD, counter, Filtering	Multiprocessors system not addressed
[3]	Modulo scheduling	DCT, FFT	Analyze one iteration, single loop
[4]	Graph conversion	Random graphs	Less parallelism, single loop
[5–7]	Loop unrolling	Random graphs, FFT, solver equalizer	Single loop unrolling
[8]	Dynamic scheduling	Fractal generation	Loop unrolling not addressed, single loop
[9, 10]	Loop fission	JPEG compression, DCT, BPIC	Loop unrolling not addressed, single loop
This work	Multiloop unrolling	Speech system image processing, N-Body	Global unrolling factors determining, coarse-grained, heterogeneous systems