COMPaS: A Pentium Pro PC-based SMP Cluster and Its Experience

نویسندگان

  • Yoshio Tanaka
  • Motohiko Matsuda
  • Makoto Ando
  • Kazuto Kubota
  • Mitsuhisa Sato
چکیده

We have built an eight node SMP cluster called COMPaS (Cluster Of Multi-Processor Systems), each node of which is a quadprocessor Pentium Pro PC. We have designed and implemented a remote memory based user-level communication layer which provides lowoverhead and high bandwidth using Myrinet. We designed a hybrid programming model in order to take advantage of locality in each SMP node. Intra-node computations utilize a multi-threaded programming style (Solaris threads) and inter-node programming is based on message passing and remote memory operations. In this paper we report on this hybrid shared memory/distributed memory programming on COMPaS and its preliminary evaluation. The performance of COMPaS is a ected by data size and access patterns, and the proportion of inter-node communication. If the data size is small enough to all t on the cache, parallel e ciency exceeds 1.0 using the hybrid programming model on COMPaS. But the performance is limited by the low memory bus bandwidth of PC-based SMP nodes for some memory intensive workloads.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

COMPaS: a PC-based SMP cluster

82 IEEE Concurrency R symmetric multiprocessor systems have become widely available, both as computational servers and as platforms for high-performance parallel computing. The same trend is found in PCs—some have also become SMPs with many CPUs in one box. Clusters of PC-based SMPs are expected to be compact, cost-effective parallel-computing platforms. At the Real World Computing Partnership,...

متن کامل

Design of OpenMP Compiler for an SMP Cluster

In this paper, we present a design of OpenMP compiler for an SMP cluster. Although clusters of SMPs are expected to be one of the cost-e ective parallel computing platforms, both of inter and intra node parallelism must be exploited to achieve high performance. These two levels of structure complicate parallel programming. The OpenMP is an emerging standard for parallel programming on shared-me...

متن کامل

SMP PCs: A Case Study on Cluster Computing

As commodity microprocessors and networks reach performance levels comparable to those used in massively parallel processors, clusters of symmetric multiprocessors are starting to be called the supercomputers of tomorrow. At the low-end of this technology are the clusters of SMP PCs, usually based on Pentium Pro or Pentium II processors. Many groups in the academia and in the industry are setti...

متن کامل

Parallelization of Saprse Cholesky Factorization on an SMP Cluster

In this paper, we present parallel implementations of the sparse Cholesky factorization kernel in the SPLASH-2 programs to evaluate performance of a Pentium Pro based SMP cluster. Solaris threads and remote memory operations are utilized for intranode parallelism and internode communications, respectively. Sparse Cholesky factorization is a typical irregular application with a high communicatio...

متن کامل

Modeling liquid crystal structures on an SMP workstation cluster∗

We examine the scalability and performance of a legacy liquid crystal code on a PC (Beowulf) cluster consisting of 16 dualprocessor Pentium III/450s. This code was originally designed for use on a Unix workstation cluster of less than 8 machines. In particular, we examine the effectiveness of using potentially more efficient techniques such as non-blocking communication calls and whether the us...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1998