Code Partitioning in Decoupled Compilers

Decoupled access/execute architectures seek to maximize performance by dividing a given program into two separate instruction streams and executing the streams on independent cooperating processors. The instruction streams consist of those instructions involved in generating memory accesses (the Access stream) and those that consume the data (the Execute stream). If the processor running the access stream is able to get ahead of the execute stream, then dynamic pre-loading of operands will occur and the penalty due to long latency operations (such as memory accesses) will be reduced or eliminated. Although these architectures have been around for many years, the performance analyses performed have been incomplete for want of a compiler. Very little has been published on how to construct a compiler for such an architecture. In this paper we describe the partitioning method employed in Daecomp, a compiler for decoupled access/execute processors.