TY - GEN
T1 - Multispeculative additive trees in high-level synthesis
AU - Del Barrio, Alberto A.
AU - Hermida, Roman
AU - Memik, Seda Ogrenci
AU - Mendias, Jose M.
AU - Molina, Maria C.
PY - 2013/10/21
Y1 - 2013/10/21
N2 - Multispeculative Functional Units (MSFUs) are arithmetic functional units that operate using several predictors for the carry signal. The carry prediction helps to shorten the critical path of the functional unit. The average performance of these units is determined by the hit rate of the prediction. In spite of utilizing more than one predictor, none or only one additional cycle is enough for producing the correct result in the majority of the cases. In this paper we present multispeculation as a way of increasing the performance of tree structures with a negligible area penalty. By judiciously introducing these structures into computation trees, it will only be necessary to predict in certain selected nodes, thus minimizing the number of operations that can potentially mispredict. Hence, the average latency will be diminished and thus performance will be increased. Our experiments show that it is possible to improve on average 24% and 38% execution time, when considering logarithmic and linear modules, respectively.
AB - Multispeculative Functional Units (MSFUs) are arithmetic functional units that operate using several predictors for the carry signal. The carry prediction helps to shorten the critical path of the functional unit. The average performance of these units is determined by the hit rate of the prediction. In spite of utilizing more than one predictor, none or only one additional cycle is enough for producing the correct result in the majority of the cases. In this paper we present multispeculation as a way of increasing the performance of tree structures with a negligible area penalty. By judiciously introducing these structures into computation trees, it will only be necessary to predict in certain selected nodes, thus minimizing the number of operations that can potentially mispredict. Hence, the average latency will be diminished and thus performance will be increased. Our experiments show that it is possible to improve on average 24% and 38% execution time, when considering logarithmic and linear modules, respectively.
KW - High-Level synthesis
KW - Operation trees
KW - Speculation
UR - http://www.scopus.com/inward/record.url?scp=84885673127&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84885673127&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84885673127
SN - 9783981537000
T3 - Proceedings -Design, Automation and Test in Europe, DATE
SP - 188
EP - 193
BT - Proceedings - Design, Automation and Test in Europe, DATE 2013
T2 - 16th Design, Automation and Test in Europe Conference and Exhibition, DATE 2013
Y2 - 18 March 2013 through 22 March 2013
ER -