Co-processor system design for fine-grain message handling in KUMP/D

I n spzte of the attractzve features of dataflow languages, they do not fit conventzonal von Neumann systems very aell. I n thzs paper, we dzscuss practical approaches to execute dataflow program mzth stock processor. A s research vehzcles, we use two types of platforms wath stock processor: completely conventzonal platforms such as engzneerzng aorkstatzons, and a partzally conventzonal platform whzch consists of an 08-the-shelf commerczal microprocessor and a speczal coprocessor. Expertmental results zndzcate that the former can be a reasonable approach thanks to the statzc analyszs? while the latter zs more powerful than the former thanks to the co-processor.

show abstract

Datarol: a parallel machine architecture for fine-grain multithreading

Amamiya

Tomiyasu²,

Kusakabe³

Proceedings. Third Working Conference on Massively Parallel Programming Models (Cat. No.97TB100228)

View full text Add to dashboard Cite

Datarol-II, which is tolerable to the latencies caused AbstractWe discuss a design principle of massively parallel distributed-memory multiprocessor architecture which solves latency problem, and present the Datarol machine architecture. Latencies, caused by remote memory access and remote procedure call, are most serious problems in massively parallel computers. In order to eliminate the processor idle times caused by these latencies, processors must perform fast context switching among fine-grain concurrent processes.First, we present a processor architecture, called Datarol-11, that promotes efficient fine-grain multithread execution by performing fast context switching among fine-grain concurrent processes. In the Datarol-II processor, an implicit register load/store mechanism is embedded in the execution pipeline in order t o reduce memory access overhead caused by context switching. In order to reduce local memory access latency, a two-level hierarchical memory system and a load control mechanism are also introduced.Then, we present a cost-effective design of the Da t arol-I1 processor, which incorporates off-the-shelf high-end microprocessor while preserving the finegrain dataflow concept. The off-the-shelf micropre cessor Pentium is used for its core processing, and a co-processor called FMP (Fine-grain Message Processor) is designed for fine grained message handling and communication controls. The co-processor FMP is desi ned on the basis of FMD (Fine-grain Message Driven? execution model, in which fine-grain multithreaded execution is driven and controlled by simple fine-grain message communications.

show abstract

Co-processor system design for fine-grain message handling in KUMP/D

Cited by 4 publications

References 12 publications

Thread Management Mechanisms of Cefos

Thread Management Mechanisms of Cefos

Executing dataflow program with stock processor

Datarol: a parallel machine architecture for fine-grain multithreading

Contact Info

Product

Resources

About