Using Triggered Operations to Offload Rendezvous Messages

نویسندگان

  • Brian W. Barrett
  • Ron Brightwell
  • Karl S. Hemmert
  • Kyle B. Wheeler
  • Keith D. Underwood
چکیده

Historically, MPI implementations have had to choose between eager messaging protocols that require buffering and rendezvous protocols that sacrifice overlap and strong independent progress in some scenarios. The typical choice is to use an eager protocol for short messages and switch to a rendezvous protocol for long messages. If overlap and progress are desired, some implementations offer the option of using a thread. We propose an approach that leverages triggered operations to implement a long message rendezvous protocol that provides strong progress guarantees. The results indicate that a triggered operation based rendezvous can achieve better overlap than a traditional rendezvous implementation and less wasted bandwidth than an eager long protocol.

برای دانلود 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

Evaluation of an Eager Protocol Optimization for MPI

Nearly all implementations of the Message Passing Interface (MPI) employ a two-level protocol for point-to-point messages. Short messages are sent eagerly to optimize for latency, and long messages are typically implemented using a rendezvous mechanism. In a rendezvous implementation, the sender must first send a request and receive an acknowledgment before the data can be transferred. While th...

متن کامل

Implications of application usage characteristics for collective communication offload

The performance of collective communication operations is known to have a significant impact on the scalability of some applications. Indeed, the global, synchronous nature of some collective operations directly implies that they will become the bottleneck when scaling to hundreds of thousands of nodes. This fact has led many researchers to try to improve the efficiency of collective operations...

متن کامل

A New DMA Registration Strategy for Pinning-Based High Performance Networks

This paper proposes a new memory registration strategy for supporting Remote DMA (RDMA) operations over pinning-based networks, as existing approaches are insufficient for efficiently implementing Global Address Space (GAS) languages. Although existing approaches often maximize bandwidth, they require levels of synchronization that discourage one-sided communication, and can have significant la...

متن کامل

cient Cache Coherence Protocols through Re nementRatan

We address the problem of developing eecient cache coherence protocols for use in distributed systems implementing distributed shared memory (DSM) using message passing. A serious drawback of traditional approaches to this problem is that the users are required to state the desired coherence protocol at the level of asynchronous message interactions involving request, acknowledge, and negative ...

متن کامل

Distributed Computing User Interfaces for 3Worlds Ecological Simulation Framework

The 3Worlds Project used distributed computng techniques to achieve decomposing single task into several subtasks, and running on several computers. In previous design, the mechanism of achieving distributon was building a Master host as the main host, using Rendezvous method(AOT wiki, 2012) to send message between Master host to several Remote hosts. However, the drawback of these methods quic...

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

برای دانلود 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

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2011