Michael, M. and Scott, M. Simple, Fast and Practical Non-Blocking and Blocking Concurrent Queue Algorithms. Technical Report TR600, University of Rochester, 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....a complete operating system kernel built without blocking synchronization. Herlihy has developed a general methodology for implementing optimistically synchronized data structures [Herlihy 1993] and other researchers have implemented and measured the performance of several such data structures [Michael and Scott 1996]. Our research explores the use of optimistic synchronization in a different context (a parallelizing compiler for irregular, object based computations) and for a different reason (to enable efficient fine grain synchronization) Advanced parallel machines such as Tera [Alverson et al. 1990] or ....

MICHAEL, M. AND SCOTT, M. 1996. Simple, fast and practical non-blocking and blocking concurrent queue algorithms. In Proceedings of the Fifteenth Annual ACM Symposium on the Principles of Distributed Computing. Philadelphia, PA.

....primitives. At least one of the first two is implemented in most commodity microprocessors. In this paper, we evaluate three non blocking and linearizable algorithms that implement concurrent FIFO queues. They are attributed to Prakash, Lee and Johnson [24] Valois [27] and Michael and Scott [21] respectively. All three algorithms use compare and swap to atomically update the queue and share a common technique for concurrent updates. The algorithms maintain pointers to the two ends of the queue. An enqueuing dequeuing process reads the pointer to the tail head of the queue respectively ....

....The non blocking algorithms perform significantly better than locks with cacheable variables, beyond the 4 processor scale. The algorithm of Michael and Scott outperforms the algorithm of Prakash, Lee and Johnson, which in turn outperforms Valois s algorithm. This result was also reported in [21] for bus based SMPs and it is fairly intuitive, given the complexity of each algorithm and its memory management scheme. The hardware counts for the best performing non blocking queue indicate that the algorithm reduces the coherence overhead to approximately one fourth of the coherence overhead ....

M. Michael and M. Scott, Simple, Fast and Practical Non-Blocking and Blocking Concurrent Queue Algorithms, Proc. 15th ACM Symp. on Principles of Distributed Computing, 1996.

....primitives. At least one of the first two is implemented in most commodity microprocessors. In this paper, we evaluate three non blocking and linearizable algorithms that implement concurrent FIFO queues. They are attributed to Prakash, Lee and Johnson [21] Valois [24] and Michael and Scott [18] respectively. All three algorithms use compare and swap to atomically update the queue and share a common technique for concurrent updates. The algorithms maintain pointers to the two ends of the queue. An enqueuing dequeuing process reads the pointer to the tail head of the queue respectively ....

....The non blocking algorithms perform significantly better than locks with cacheable variables, beyond the 4 processor scale. The algorithm of Michael and Scott outperforms the algorithm of Prakash, Lee and Johnson, which in turn outperforms Valois s algorithm. This result was also reported in [18] for bus based SMPs and it is fairly intuitive, given the complexity of each algorithm and its memory management scheme. The hardware counts for the best performing non blocking queue indicate that the algorithm reduces the coherence overhead to approximately one fourth of the coherence overhead ....

M. Michael and M. Scott, Simple, Fast and Practical NonBlocking and Blocking Concurrent Queue Algorithms, Proc. 15th ACM Symp. on Principles of Distributed Computing, 1996.

....kernel [20] a complete operating system kernel built without blocking synchronization. Herlihy has developed a general methodology for implementing optimistically synchronized data structures [12] and other researchers have implemented and measured the performance of several such data structures [23]. Our research explores the use of optimistic synchronization in a different context (a parallelizing compiler for irregular, object based computations) and for a different reason (to enable efficient finegrain synchronization) Several software systems use the state bits and synchronizing read ....

M. Michael and M. Scott. Simple, fast and practical non-blocking and blocking concurrent queue algorithms. In Proceedings of the Fifteenth Annual ACM Symposium on the Principles of Distributed Computing, Philadelphia, PA, May 1996.

No context found.

Michael, M. and Scott, M. Simple, Fast and Practical Non-Blocking and Blocking Concurrent Queue Algorithms. Technical Report TR600, University of Rochester, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC