This paper aims to implement high performance OOPL on conventional multicomputers.
Highly-tuned Hardware & Specially tailored Software
Two key factors in high-performance concurrent OOPL and their hardware assisted solutions:
- Efficient Message Passing: processors and the network are tightly connected.
- Efficient intra-node multithreading: the hardware manages the scheduling queue and the next runnable thread is scheduled automatically upon termination of the current thread.
Stack/LIFO-based scheduling - Intra-node Communication