TY - JOUR
T1 - Design exploration framework under impreciseness based on register-constrained inclusion scheduling
AU - Chantrapornchai, Chantana
AU - Surakumpolthorn, Wanlop
AU - Sha, Edwin
PY - 2004
Y1 - 2004
N2 - In this paper, we propose a design exploration framework which consider impreciseness in design specification. In high-level synthesis, imprecise information is often encountered. Two kinds of impreciseness are considered here: imprecise characteristics of functional units and imprecise design constraints. The proposed design exploration framework is based on efficient scheduling algorithm which considers impreciseness, Register-Constrained Inclusion Scheduling. We demonstrate the effectiveness of our framework by exploring a design solution for a well-known benchmark. Voltera filter. The selected solution meets the acceptability criteria while minimizing the total number of registers.
AB - In this paper, we propose a design exploration framework which consider impreciseness in design specification. In high-level synthesis, imprecise information is often encountered. Two kinds of impreciseness are considered here: imprecise characteristics of functional units and imprecise design constraints. The proposed design exploration framework is based on efficient scheduling algorithm which considers impreciseness, Register-Constrained Inclusion Scheduling. We demonstrate the effectiveness of our framework by exploring a design solution for a well-known benchmark. Voltera filter. The selected solution meets the acceptability criteria while minimizing the total number of registers.
KW - Imprecise Design Exploration
KW - Imprecise information
KW - Inclusion Scheduling
KW - Multiple design attributes
KW - Register constraint
KW - Scheduling/Allocation
UR - https://www.scopus.com/pages/publications/35048824373
U2 - 10.1007/978-3-540-30502-6_6
DO - 10.1007/978-3-540-30502-6_6
M3 - 文章
AN - SCOPUS:35048824373
SN - 0302-9743
VL - 3321
SP - 78
EP - 92
JO - Lecture Notes in Computer Science
JF - Lecture Notes in Computer Science
ER -