Efficacy and performance impact of value prediction
Bohuslav Rychlik, John Faistl, Bryon Krug, John Paul Shen · 1998
this paper are generated by an execution-driven performance simulator [2]. The simulator uses a cycle-accurate machine model based on the PowerPC 604 [3][6][17]. All key aspects of the PowerPC 604 microarchitecture are modeled. The SPEC95 benchmark set is used for this work, including eight integer and six floating point programs. The benchmark programs, input sets, and run lengths are summarized in Table 1. To reduce simulation times, we use small input files and limit benchmark length to 100 million instructions. All user library calls are modeled, though system calls are not.