Communion: a new strategy for memory management in high-performance computer systems

• Autores: Midorikawa, Edson T.; Sato, Liria Matsumoto; Zuffo, João Antônio
• Año de publicación: 1997
• Idioma: inglés
• Tipo de recurso: documento de conferencia
• Materia: Ciencias Informáticas; Parallel processing; Distributed; computer systems; new strategy; memory management
• Institución: Universidad Nacional de La Plata

Modern computers present a big gap between peak performance and sustained performance. There are many reasons for this situation, but mainly involving an inefficient usage of computational resources. Nowadays the memory system is the most critical component because of its growing inability to keep up with the processor requests. Technological trends have produced a large and growing gap between CPU speeds and DRAM speeds. Much research has focused this memory system problem, including program optimizing techniques, data locality enhancement, hardware and software prefetching, decoupled architectures, mutithreading, speculative loads and execution. These techniques have got a relative success, but they focus only one component in the hardware or software systems. We present here a new strategy for memory management in high-performance computer systems, named COMMUNION. The basic idea behind this strategy is cooperation. We introduce some interaction possibilities among system programs that are responsible to generate and execute application programs. So, we investigate two specific interactions: between the compiler and the operating system, and among the compiling system components. The experimental results show that it’s possible to get improvements of about 10 times in execution time, and about 5 times in memory demand. In the interaction between compiler and operating system, named Compiler-Aided Page Replacement (CAPR), we achieved a reduction of about 10% in space-time product, with an increase of only 0.5% in the total execution time. All these results show that it’s possible to manage main memory with a better efficiency than current systems.
Eje: Procesamiento distribuido y paralelo. Tratamiento de señales
Red de Universidades con Carreras en Informática (RedUNCI)