An-Najah Staff

In this paper we present the Data-Driven Multithreading Virtual Machine for the Cell Processor (DDM-VM ). DataDriven Multithreading is a non-blocking multithreading model that decouples the synchronization from the computation portions of a program allowing them to execute asynchronously in a data°ow manner. The core of the DDM model is the Thread Scheduling Unit (TSU) which schedules threads dynamically at runtime based on data availability. DDM-VM implements the TSU as a software module running on the PPE core of the Cell, allowing the SPE cores to execute the program threads. DDM-VM virtualizes the parallel resources of the Cell processor, handles the heterogeneity of the cores and manages the Cell memory hierarchy e±ciently. c We present the architecture of DDM-VM and provide an in-depth performance analysis using a suite of standard computational benchmarks. The evaluation shows that DDMVM  scales well and tolerates scheduling overheads and memory latencies e®ectively. Furthermore, DDM-VM compares favorably with other platforms targeting the Cell processor.