Abstract
The SMiLE project (Shared Memory in a LAN-like Environment) at LRR-TUM investigates high-performance cluster computing based on SCI interconnect technology [9]. The major research goal is to develop concepts, programming models, and tools for the efficient use of a distributed shared memory (DSM) system with NUMA (non-uniform memory access) characteristics. One focus within the project is to adapt and optimize “standard” parallel processing software for SCI-based cluster environments. Several communication libraries [8], a POSIX compliant distributed thread package [17], and the MuSE runtime system combining DSM, multi-threading and data-flow techniques [13] have been developed and implemented making use of SCI’s hardware-based DSM.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
