WildFire: a scalable path for SMPs

Hägersten, Erik; Köster, Michael

doi:10.1109/hpca.1999.744361

Cited by 84 publications

(81 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Two implementations of dynamic page migration on real systems (the SGI Origin2000 [15] and the Sun Wildfire [9,23]) have not demonstrated any noteworthy performance improvements [11]. The reasons for this are the limited accuracy of the page migration algorithms used in the implementations, the sensitivity to transient effects in the workloads and the inability to amortize the cost of page migrations [21].…”

Section: Memory Management Techniquesmentioning

confidence: 99%

“…In addition to this service, IRIX has a complete memory management control interface exported to the user. Although both features are platform-specific, similar services are provided in other commercial cache-coherent DSMs, such as the Sun Wildfire [9,23]. Furthermore, both commodity and open source operating systems such as TruUNIX and implementations of Linux backed up by SGI, IBM, HP and Sun [16] are currently incorporating features to support memory management services for NUMA systems, such as locality-based thread scheduling and demand-based page placement and migration.…”

Section: User-level Implementationmentioning

confidence: 99%

See 1 more Smart Citation

Scheduler-Activated Dynamic Page Migration for Multiprogrammed DSM Multiprocessors

Nikolopoulos

Polychronopoulos

Papatheodorou

et al. 2002

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

The performance of multiprogrammed shared-memory multiprocessors suffers often from scheduler interventions that neglect data locality. On cachecoherent distributed shared-memory (DSM) multiprocessors, such scheduler interventions tend to increase the rate of remote memory accesses. This paper presents a novel dynamic page migration algorithm that remedies this problem in iterative parallel programs. The purpose of the algorithm is the early detection of pages that will most likely be accessed remotely by threads associated with them via a thread-to-memory affinity relation. The key mechanism that enables timely identification of these pages is a communication interface between the page migration engine and the operating system scheduler. The algorithm improves previously proposed competitive page migration algorithms in many aspects, including accuracy, timeliness and cost amortization. Most notably, the algorithm is not biased by obsolete memory access history that may be accumulated in the page access counters at runtime. Experiments on the SGI Origin2000 show that the algorithm outperforms by far the best static page placement algorithm and a customized page migration engine implemented in IRIX, the Origin2000 operating system. The algorithm is implemented at user-level and its functionality is orthogonal to the scheduling policy of the operating system. # 2002 Elsevier Science (USA)

show abstract

Section: Memory Management Techniquesmentioning

confidence: 99%

Section: User-level Implementationmentioning

confidence: 99%

Scheduler-Activated Dynamic Page Migration for Multiprogrammed DSM Multiprocessors

Nikolopoulos

Polychronopoulos

Papatheodorou

et al. 2002

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

show abstract

“…The hardware DSM numbers have been measured on a 2-node Sun Orange built from two E6000 nodes connected through a hardware-coherent interface with a raw bandwidth of 800 Mbyte/s in each direction [15], [30]. The Orange system has been configured as a traditional cache-coherent, non-uniform memory access (CC-NUMA) architecture with its data migration capability activated while its coherent memory replication (CMR) has been kept inactive.…”

Section: Methodsmentioning

confidence: 99%

“…Blocking directory coherence protocols have been suggested to simplify the design and verification of hardware DSM systems [15]. The directory blocks new requests to a cache line until all previous coherence activity to the cache line has ceased.…”

Section: Blocking Directory Protocol Overviewmentioning

confidence: 99%

See 1 more Smart Citation

Removing the overhead from software-based shared memory

Radović

Hägersten

2001

Proceedings of the 2001 ACM/IEEE Conference on Supercomputing

View full text Add to dashboard Cite

show abstract

Performance of High-Accuracy PDE Solvers on a Self-Optimizing NUMA Architecture

Holmgren

Wallin

2001

Euro-Par 2001 Parallel Processing

View full text Add to dashboard Cite

Abstract. High-accuracy PDE solvers use multi-dimensional fast Fourier transforms. The FFTs exhibits a static and structured memory access pattern which results in a large amount of communication. Performance analysis of a non-trivial kernel representing a PDE solution algorithm has been carried out on a Sun WildFire computer. Here, different architecture, system and programming models can be studied. The WildFire system uses self-optimization techniques such as data migration and replication to change the placement of data at runtime. If the data placement is not optimal, the initial performance is degraded. However, after a few iterations the page migration daemon is able to modify the placement of data. The performance is improved, and equals what is achieved if the data is optimally placed at the start of the execution using hand tuning. The speedup for the PDE solution kernel is surprisingly good.

show abstract

WildFire: a scalable path for SMPs

Cited by 84 publications

References 14 publications

Scheduler-Activated Dynamic Page Migration for Multiprogrammed DSM Multiprocessors

Scheduler-Activated Dynamic Page Migration for Multiprogrammed DSM Multiprocessors

Removing the overhead from software-based shared memory

Performance of High-Accuracy PDE Solvers on a Self-Optimizing NUMA Architecture

Contact Info

Product

Resources

About