Virtual Interface Architecture Draft Write-up
نویسندگان
چکیده
INTRODUCTION The recent arrivals of fast networks, such as Myrinet, asynchronous transfer mode (ATM) coupled with powerful personal computers and workstations, has allowed for cost effective high-performance and high-availability clusters to be built from commodity parts. Additionally, the new cluster platform offers enough raw hardware performance to allow execution of applications such as parallel scientific computing or parallel servers, which used to be run on massively parallelised processors (MPPs) and symmetric multi-processors (SMPs). However, to achieve good results, the cluster communication subsystem must deliver user-to-user performance at least comparable to that of MPPs.
منابع مشابه
Reorder the Write Sequence by Virtual Write Buffer to Extend SSD's Lifespan
The limited lifespan is the Achilles’s heel of Solid State Drive (SSD) based on NAND flash memory. NAND flash has two drawbacks that degrade SSD’s lifespan. One is the out-of-place update. Another is the sequential write constraint within a block. To extend the lifespan, SSD usually employs a write buffer to reduce write traffic to flash memory. However, existing write buffer schemes only pay a...
متن کاملP8P Parallel Phase Change Memory (PCM) data sheet
Features • High-performance READ – 115ns initial READ access – 135ns initial READ access – 25ns, 8-word asynchronous-page READ • Architecture – Asymmetrically blocked architecture – Four 32KB parameter blocks with top or bottom configuration – 128KB main blocks – Serial peripheral interface (SPI) to enable lower pin count on-board programming • Phase change memory (PCM) – Chalcogenide phase cha...
متن کاملThe Region Trap Library: Handling Traps on Application-Defined Regions of Memory
User-level virtual memory (VM) primitives are used in many different application domains including distributed shared memory, persistent objects, garbage collection, and checkpointing. Unfortunately, VM primitives only allow traps to be handled at the granularity of fixedsized pages defined by the operating system and architecture. In many cases, this results in a size mismatch between pages an...
متن کاملA Framework for Managing Multicore Resources
The computer industry is undergoing a momentous transformation. General-purpose computing is moving off of desktops and onto diverse device such as smart phones, digital entertainment centers, and data center servers. At the same time, high-performance semiconductor manufacturers are continuing the long-term trend of increasing integration and are now manufacturing multicore chips. It is our po...
متن کاملAbout the Performance of HPF: Improving Runtime on the Cray T3E with Hardware Specific Properties
High Performance Fortran permits to write parallel programs with much less programming effort than by using standard communication libraries such as MPI or PVM. The performance of compiled HPF programs is considered low, though. We show that a compiled HPF application will gain a substantial runtime improvement if compilation incorporates properties of the hardware architecture into the final p...
متن کامل