Filters








23,997 Hits in 3.6 sec

Destination-passing style for efficient memory management

Amir Shaikhha, Andrew Fitzgibbon, Simon Peyton Jones, Dimitrios Vytiniotis
2017 Proceedings of the 6th ACM SIGPLAN International Workshop on Functional High-Performance Computing - FHPC 2017  
The key idea is to transform the program to destination passing style, which in turn enables a highly-efficient stack-like memory allocation discipline.  ...  In this paper we propose a technique that automates a common custom memory-management technique, which we call destination passing style (DPS), as used in efficient C and Fortran libraries such as BLAS  ...  This gives the programmer full control over memory management for string values. Other languages have exploited destination-passing style during compilation [15, 16] .  ... 
doi:10.1145/3122948.3122949 dblp:conf/icfp/ShaikhhaFJV17 fatcat:agqiyqo3lbevrapjkaeuldgobq

Distributed Shared Memory Based Live VM Migration

Tariq Daradkeh, Anjali Agarwal
2016 2016 IEEE 9th International Conference on Cloud Computing (CLOUD)  
Live virtual machine migration is an essential tool for dynamic resource management in current data centers.  ...  Finally in the resume and commitment phase, the migrated VM is resumed running in the destination physical host, the remaining memory pages are pulled by destination machine from the source machine.  ...  One is called message passing style of explicitly parallelized program and the other is shared memory space parallelism.  ... 
doi:10.1109/cloud.2016.0116 dblp:conf/IEEEcloud/DaradkehA16 fatcat:hedwazqzfffbhmar6pcta6rqz4

Supporting systolic and memory communication in iWarp

Shekhar Borkar, Craig Peterson, Jim Susman, Jim Sutton, John Urbanski, Jon Webb, Robert Cohn, George Cox, Thomas Gross, H. T. Kung, Monica Lam, Margie Levine (+2 others)
1990 Proceedings of the 17th annual international symposium on Computer Architecture - ISCA '90  
The former allows programs to access data as they are transmitted and to redirect portions of messages to different destinations efficiently.  ...  Therefore communication begins or terminates at the local memory. For conventional message passing methods, both sending and receiving processors use memory communication.  ...  We also thank Abu Noamsn and David Yam of Carnegie Mellon University for assistance in design validation and performance evaluation.  ... 
doi:10.1145/325164.325116 dblp:conf/isca/BorkarCCGKLLMMPSSUW90 fatcat:u34kdq5fdjei5ngq4vop4h34hq

Decoupling synchronization and data transfer in message passing systems of parallel computers

T. Stricker, J. Stichnoth, D. O'Hallaron, S. Hinrichs, T. Gross
1995 Proceedings of the 9th international conference on Supercomputing - ICS '95  
be deposited directly into the final destination in the memory of the remote recipient.  ...  Measurements of these three styles on a Cray T3D demonstrate the benefits of the decoupled message passing with direct deposit.  ...  In our decoupled deposit message passing system, all messages are taken directly from memory (user space) at the sender and are automatically directed to their final destination in memory at the receiving  ... 
doi:10.1145/224538.224539 dblp:conf/ics/StrickerSOHG95 fatcat:qgnr6mz5bjf3tbzxcibsoc7y24

A routerless system level interconnection network for 3D integrated systems

Kelli Ireland, Donald Chiarulli, Steven Levitan
2009 2009 IEEE International Conference on 3D System Integration  
These models, rather than message passing, have the easiest path for migration of applications from current sequential models.  ...  The architecture is scalable enough to meet the needs of next generation multicore systems and can efficiently support multiple programming models including symmetric common memory architectures.  ...  This design supports efficient and equitable management of the fan-in networks within each partition. It operates transparently, with minimal control latency.  ... 
doi:10.1109/3dic.2009.5306557 dblp:conf/3dic/IrelandCL09 fatcat:b5lzksa7gjhdvfjnyehvdutqee

Integrating message-passing and shared-memory

David Kranz, Kirk Johnson, Anant Agarwal, John Kubiatowicz, Beng-Hong Lim
1993 SIGPLAN notices  
Because message-passing mechanisms are much more efficient than shared-memory loads and stores for certain types of interprocessor communication and synchronization operations, however, we argue for building  ...  multiprocessors that efficiently support both shared-memory and message-passing mechanisms.  ...  We would also like to thank Alan Mainwaring, Dave Douglas, and Thinking Machines Corporation for their generosity and assistance in porting our simulation system to the CM-5.  ... 
doi:10.1145/173284.155338 fatcat:dipn6mfrc5aupi2cwgy7jxdbdy

Integrating message-passing and shared-memory

David Kranz, Beng-Hong Lim, Kirk Johnson, John Kubiatowicz, Anant Agarwal
1993 SIGPLAN notices  
Because message-passing mechanisms are much more efficient than shared-memory loads and stores for certain types of interprocessor communication and synchronization operations, however, we argue for building  ...  multiprocessors that efficiently support both shared-memory and message-passing mechanisms.  ...  We would also like to thank Alan Mainwaring, Dave Douglas, and Thinking Machines Corporation for their generosity and assistance in porting our simulation system to the CM-5.  ... 
doi:10.1145/156668.156705 fatcat:zetpzu4x45g37nx7y6xyruli5m

Integrating message-passing and shared-memory

David Kranz, Kirk Johnson, Anant Agarwal, John Kubiatowicz, Beng-Hong Lim
1993 Proceedings of the fourth ACM SIGPLAN symposium on Principles and practice of parallel programming - PPOPP '93  
Because message-passing mechanisms are much more efficient than shared-memory loads and stores for certain types of interprocessor communication and synchronization operations, however, we argue for building  ...  multiprocessors that efficiently support both shared-memory and message-passing mechanisms.  ...  We would also like to thank Alan Mainwaring, Dave Douglas, and Thinking Machines Corporation for their generosity and assistance in porting our simulation system to the CM-5.  ... 
doi:10.1145/155332.155338 dblp:conf/ppopp/KranzJAKL93 fatcat:6tjf3vkdyzgdxdkri4k6pfuw74

Anatomy of a message in the Alewife multiprocessor

John Kubiatowicz, Anant Agarwal
2014 25th Anniversary International Conference on Supercomputing Anniversary Volume -  
For efficiency, it is often implemented with a layer of interpretive hardware on top of a message-passing communications network.  ...  Accordingly, Alewife provides support for both styles of communication.  ...  For example, the iWarp [9] integrates systolic and message passing styles of communications.  ... 
doi:10.1145/2591635.2667168 fatcat:tb34jzztxfbbre5jcxvpnlaaba

Anatomy of a message in the Alewife multiprocessor

John Kubiatowicz, Anant Agarwal
1993 Proceedings of the 7th international conference on Supercomputing - ICS '93  
For efficiency, it is often implemented with a layer of interpretive hardware on top of a message-passing communications network.  ...  Accordingly, Alewife provides support for both styles of communication.  ...  For example, the iWarp [9] integrates systolic and message passing styles of communications.  ... 
doi:10.1145/165939.165970 dblp:conf/ics/KubiatowiczA93 fatcat:mtr6rwmanrfr3bc3rgsyimzjle

Interprocess communications in the AN/BSY-2 distributed computer system: a case study

David Andrews, Paul Austin, Peter Costello, David LeVan
2002 Journal of Systems and Software  
The message-passing semantics are then presented. A key feature of the IPC facility is it's support for both synchronous and asynchronous communications based on logical addressing.  ...  This paper presents a case study of the design and implementation of the interprocess communications facility developed for the AN/BSY-2 distributed computer system, the computer system for the Seawolf  ...  As implementation of message passing facilities became more efficient, its popularity as a scalable communications model continued to grow.  ... 
doi:10.1016/s0164-1212(01)00151-0 fatcat:24j363nk4nctpgjqzkbkegaety

A case study of a system level approach to exploration of queuing management schemes for input queue packet switches

Chen He, M. Lajolo, M. Jacome
2003 Eleventh Euromicro Conference on Parallel, Distributed and Network-Based Processing, 2003. Proceedings.  
The effectiveness of this novel method is demonstrated by a case study of exploring data queuing management schemes for packet switches with input queues.  ...  queuing management system.  ...  Figure 8 shows the effect of different packet sizes on the energy efficiency for both memory management schemes under both uniform (left) and non-uniform (right) traffic.  ... 
doi:10.1109/empdp.2003.1183617 dblp:conf/pdp/HeLJ03 fatcat:pkcdyeemjrac3awp6bsgnunqyq

The Efficiency of Linda for General Purpose Scientific Programming

Timothy G. Mattson
1994 Scientific Programming  
It is easy to use, and lets the programmer code in a very expressive, uncoupled programming style. These benefits, however, are of little value unless Linda programs execute efficiently.  ...  The Linda programming model is based on operations that read, write, and erase a virtual shared memory.  ...  Hence, no extra messages are required to manage tuple space causing the Linda and the message passing versions of the programs to pass essentially the same number of messages.  ... 
doi:10.1155/1994/401086 fatcat:sl4k7i7d7jghlmf2kpkfardsqy

BOLT: Energy-efficient Out-of-Order Latency-Tolerant execution

Andrew Hilton, Amir Roth
2010 HPCA - 16 2010 The Sixteenth International Symposium on High-Performance Computer Architecture  
LT (latency tolerant) execution is an attractive candidate technique for future out-of-order cores.  ...  Collectively, these modifications turn outof-order LT into a technique that improves performance in an energy-efficient way.  ...  Acknowledgments We thank the reviewers for their comments. This work was supported by NSF grant CCF-0541292.  ... 
doi:10.1109/hpca.2010.5416634 dblp:conf/hpca/HiltonR10 fatcat:eegugksit5alzh2wx5tqg3dqcy

Zippy: A Framework for Computation and Visualization on a GPU Cluster

Zhe Fan, Feng Qiu, Arie E. Kaufman
2008 Computer graphics forum (Print)  
Zippy preserves the advantages of both message passing and shared-memory models. It employs global arrays (GA) to simplify the communication, synchronization, and collaboration among multiple GPUs.  ...  It abstracts the GPU cluster programming with a two-level parallelism hierarchy and a non-uniform memory access (NUMA) model.  ...  Acknowledgements We wish to thank Jarek Nieplocha and Manojkumar Krishnan for discussing GA and Mike Houston for discussing parallel volume rendering. This work is supported by NSF grant CCF-0702699.  ... 
doi:10.1111/j.1467-8659.2008.01131.x fatcat:bchxru3j6ffdli36u57rjirhsm
« Previous Showing results 1 — 15 out of 23,997 results