Results 1 -
9 of
9
Home Migration in Home-Based Software DSMs
- IN FIRST WORKSHOP ON SOFTWARE DISTRIBUTED SHARED MEMORY
, 1999
"... Home-based software DSMs provide a simple, effective, and scalable way to build software DSMs. However, the performance of home-based software DSMs is sensitive to the distribution of home pages. This paper introduces our work on migrating home pages adaptively according to the application sharing p ..."
Abstract
-
Cited by 16 (6 self)
- Add to MetaCart
Home-based software DSMs provide a simple, effective, and scalable way to build software DSMs. However, the performance of home-based software DSMs is sensitive to the distribution of home pages. This paper introduces our work on migrating home pages adaptively according to the application sharing pattern in a home-based software DSM system called JIAJIA. In the scheme, pages that are written by only one processor between two barriers are migrated to the single writing processor. Migration messages are piggybacked on barrier messages and no additional communication is required for the migration. Though very simple, performance evaluation with SPLASH program suite and NAS Parallel Benchmarks shows that home migration can reduce diffs dramatically and performance gains obtained by home migration arranges from several to hundreds percent compared to statically distributing home of shared data page-by-page across processors.
Reducing Message Overhead in Home-Based Software DSMs
- IN PROC. OF THE 1ST WORKSHOP ON SOFTWARE DISTRIBUTED SHARED MEMORY
, 1999
"... Among the tradeoffs between home-based and homeless software DSMs, a frequently mentioned disadvantage of the home-based software DSM is that it fetches the whole page on a page fault while the homeless protocol fetches diffs only. This paper proposes the write vector technique to reduce message amo ..."
Abstract
-
Cited by 5 (2 self)
- Add to MetaCart
Among the tradeoffs between home-based and homeless software DSMs, a frequently mentioned disadvantage of the home-based software DSM is that it fetches the whole page on a page fault while the homeless protocol fetches diffs only. This paper proposes the write vector technique to reduce message amount in home-based software DSMs. Other than fetching a whole page on a page fault as in traditional home-based software DSMs, the write vector technique divides a page into blocks and fetches only those blocks that are dirty with respect to the faulting processor on a page fault. A write vector table is maintained for each shared page in its home to record for each processor which block(s) has been modified since the processor fetched the page last time. Performance evaluation with some popularly accepted benchmarks shows that the write vector technique can reduce message amounts dramatically and consequently improve performance significantly in some benchmarks. The extra time overhead of th...
Affinity-based Self Scheduling: A More Practical Load Balancing Scheme for Home-based Software DSMs
, 1999
"... Load balancing is a critical issue for achieving good performance in any parallel and distributed systems. However, this issue is neglected in the research area of software DSM systems in the past twelve years, where all the performance evaluation of previous work are based on dedicated homogeneous ..."
Abstract
-
Cited by 3 (2 self)
- Add to MetaCart
Load balancing is a critical issue for achieving good performance in any parallel and distributed systems. However, this issue is neglected in the research area of software DSM systems in the past twelve years, where all the performance evaluation of previous work are based on dedicated homogeneous hardware environment, such as shared memory multiprocessors, distributed memory systems. The main contribution of this paper includes two aspects: Firstly, to our knowledge, this is the first paper that evaluate and compare several well known different scheduling schemes in a software DSM system with metacomputing environment. We find that many theory analytical results in the literature does not work any more. Secondly, based on our analysis, we present and evaluate a new affinity-based (ABS) dynamic self scheduling algorithm for home-based software DSM system. In this scheme, the static affinity between processor and initial data distribution is taken into account when schedulin...
Optimizing Home-based Software DSM Protocols, Cluster Computing:
- The Journal of Networks, Software and Applications, Baltzer Science Publishers,
, 2001
"... ..."
(Show Context)
Running Real Applications on Software DSMs
- Proc. of 4th Int. Conf. on High Performance Computing in Asia-Pacific Region
, 2000
"... This paper introduces our experiences with some real applications on the home-based software DSM JIAJIA and discusses techniques of parallelizing a sequential program to run on software DSM. It categorizes parallel program segments into five patterns: single-process sequential, mutual-exclusive sequ ..."
Abstract
-
Cited by 1 (1 self)
- Add to MetaCart
(Show Context)
This paper introduces our experiences with some real applications on the home-based software DSM JIAJIA and discusses techniques of parallelizing a sequential program to run on software DSM. It categorizes parallel program segments into five patterns: single-process sequential, mutual-exclusive sequential, data-parallel, task-parallel, and common-parallel. The usage of each pattern is then discussed with the real applications as examples. With some guide from their owners, these programs are parallelized to the API of JIAJIA in a very short time. Satisfactory speedups are achieved for them on a cluster of eight Pentium II PCs connected by a 100Mbps switched Ethernet. Our experiences imply that with the advances of software DSMs and network technologies, the time for pushing the software DSM into the parallel processing mainstream has come and efforts should be made by software DSM researchers to expand the application of software DSMs. 1 Introduction Over the past decade, software Di...
© 2001 Kluwer Academic Publishers. Manufactured in The Netherlands. Optimizing Home-Based Software DSM Protocols ∗
"... Abstract. Software DSMs can be categorized into homeless and home-based systems both have strengths and weaknesses when compared to each other. This paper introduces optimization methods to exploit advantages and offset disadvantages of the home-based protocol in the home-based software DSM JIAJIA. ..."
Abstract
- Add to MetaCart
(Show Context)
Abstract. Software DSMs can be categorized into homeless and home-based systems both have strengths and weaknesses when compared to each other. This paper introduces optimization methods to exploit advantages and offset disadvantages of the home-based protocol in the home-based software DSM JIAJIA. The first optimization reduces the overhead of writes to home pages through a lazy home page write detection scheme. The normal write detection scheme write-protects shared pages at the beginning of a synchronization interval, while the lazy home page write detection delays home page write-protecting until the page is first fetched in the interval so that home pages that are not cached by remote processors do not need to be write-protected. The second optimization avoids fetching the whole page on a page fault through dividing a page into blocks and fetching only those blocks that are dirty with respect to the faulting processor. A write vector table is maintained for each shared page in its home to record for each processor which block(s) has been modified since the processor fetched the page last time. The third optimization adaptively migrates home of a page to the processor most frequently writes to the page to reduce twin and diff overhead. Migration information is piggybacked on barrier messages and no additional communication is required for the migration. Performance evaluation with some well-accepted benchmarks and real applications shows that the above optimization methods can reduce page faults, message amounts, and diffs dramatically and consequently improve performance significantly.
Where Does the Time Go in Software DSM Systems: Experiences with JIAJIA?
, 1999
"... The performance gap between software DSM systems and message passing platforms prevents the prevalence of software DSM system greatly, though great efforts have been delivered in this area in the past decade. In this paper, we take the challenge to find where should we focus our strength on in the f ..."
Abstract
- Add to MetaCart
The performance gap between software DSM systems and message passing platforms prevents the prevalence of software DSM system greatly, though great efforts have been delivered in this area in the past decade. In this paper, we take the challenge to find where should we focus our strength on in the future design. The components of total system overhead of software DSM systems are analyzed in detail firstly. Based on a state-ofthe -art software DSM system JIAJIA, we measure these components on Dawning parallel system and draw five important conclusions which are different from some traditional viewpoints. (1) The performance of the JIAJIA software DSM system is acceptable. For four of eight applications, the parallel efficiency achieved by JIAJIA is about 80%, while for two others, 70% efficiency can be obtained. (2) 40.94% interrupt service time is overlapped with waiting time. (3) Encoding and decoding diffs do not cost much time(!1%), so using hardware support to encode/dec...
FAST PARALLEL IMAGE MATCHING ALGORITHM ON CLUSTER
"... This paper introduces our experiences with developing fast Parallel stereo matching algorithm on Cluster. The key technique of generating DEMs or DSMs ’ from remote-sensing images is stereo matching, and it’s one of the most time consuming algorithms, so that many efforts have been carried out to de ..."
Abstract
- Add to MetaCart
(Show Context)
This paper introduces our experiences with developing fast Parallel stereo matching algorithm on Cluster. The key technique of generating DEMs or DSMs ’ from remote-sensing images is stereo matching, and it’s one of the most time consuming algorithms, so that many efforts have been carried out to develop fast algorithms. Many difficulty problems continue to challenge researchers in this field; they are occlusion, large parallax range, and radiant distortion. To solve these problems need more complicated imagematching algorithm that is more time consuming. Recent advance in high-speed networks, rapid improvement in microprocessor design, and availability of highly performing clustering software implementations enables cost-effective high-performance parallel computing on clusters very attractive. The cluster is a very useful platform on which to develop fast parallel image matching algorithm dealing with the difficult problems above-mentioned. A novel operator of image matching based on no-uniform image resample is used in parallel image matching algorithm, this operator reduces the influence of geometric distortion on those matching operators which are in effect under equal parallax assumption, for example, correlation operator, and expands the scope of them to be in effect. 1.
