Toward more advanced usage of instruction level parallelism by a very large data path processor architecture
H. Tanaka · 2002
The architectural performance gain of a microprocessor is going to saturate because of the small gain of instruction level parallelism. In this paper, we discuss the design points and some tentative solutions to overcome this bottleneck and propose a processor architecture called Very Large Data Path. This architecture broadens the window of instruction analysis to extract 10 times of parallel gain compared with the conventional superscaler processors. This paper discusses the system elements and shows some preliminary evaluation results.