PERFORMANCE OF LEVEL-3 BLAS KERNELS IN A DYNAMICALLY PARTITIONED DATA-FLOW ENVIRONMENT

The Dynamically Partitioned Data-Flow (DPDF) model is based on an orig inal analysis concept of the data dependency graph at the instruction level. Instead of a breadth first analysis, as in a classical Data-Flo w Model, we execute instructions along data-dependent paths. As a cons equence, data locality can be exploited by reusing results between the execution of consecutive instructions. In addition, the different pat hs are not statically defined but arise from a dynamical partitioning of the graph. This model presents the advantage to support very small cost dynamic scheduling and multitasking strategies. In order to study the efficiency of this new model, a first architecture has been defin ed. This architecture is currently limited to a single processor with one serial processing unit but four graph analyzing units (called pref etch units). Each of these prefetch units is able to build dynamically its own execution path inside the Data-Flow graph of an application. The efficiency of this architecture is studied on a numerical benchmar k composed of a subset of the Livermore loops and of three routines of the Level 3 BLAS (GEMM, SYRK and TRSM). Our goal in these experimenta tions is to demonstrate the ability of the four prefetch units to feed the ALU.