Windows NT in a ccNUMA system

B. C. Brock, G. D. Carpenter, E. Chiprout, E Elnozahy, M. E. Dean, D. Glasco, J. Peterson, Ram Rajamony, F. L. Rawson, Ron Rockhold, A Zimmerman

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

We have built a 16-way. ccNUMA multiprocessor prototype to study the feasibility of building large scale servers out of Standard High Volume (SHV) components. Using a cache-coherent interconnect, our prototype combines four 4-processor SMPs built using 350MHz Intel Xeon(TM) processors, yielding a 16-way system with a total of 4 GBytes of physical memory distributed over the nodes. Such an environment poses several performance challenges to Windows NT(R). which assumes that memory is equidistant to all processors. To overcome these problems, we have implemented an abstraction called a Resource Set which allows threads to specify their execution and memory affinity across the ccNUMA complex.We used a suite of parallel applications to evaluate the scalability and performance of the system. Our results confirm the feasibility of building ccNUMA systems out of SHV components, and suggest that memory allocation affinity should be incorporated as part of the standard Windows NT API. Also, the performance degradation due to poor bus bandwidth in the current generation of Intel-based processors often dominates the degradation due to the latency of remote memory accesses.
Original languageEnglish (US)
Title of host publicationPROCEEDINGS OF THE 3RD USENIX WINDOWS NT SYMPOSIUM
PublisherUSENIX ASSOC
Pages61-72
Number of pages12
ISBN (Print)1-880446-29-4
StatePublished - 1999
Externally publishedYes

Fingerprint Dive into the research topics of 'Windows NT in a ccNUMA system'. Together they form a unique fingerprint.

Cite this