Efficient all-to-all broadcast schemes in distributed-memory parallel computers

Oh, Eunseuk; Kanj, Iyad A.

doi:10.1109/hpcsa.2002.1019136

Cited by 3 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many all-to-all broadcast algorithms were designed for specific network topologies that are used in parallel machines, including hypercube (3,4) , mesh (5)(6)(7) , torus (7) , k-ary n-cube (4) , fat tree (8) , and star (9) . Work in (10) optimizes MPI collective communications, including MP I All gather, on wide area networks.…”

Section: Related Workmentioning

confidence: 99%

Bandwidth Efficient All-to-All Broadcast on Switched Clusters

Faraj

Patarasuk

Yuan

2007

Int J Parallel Prog

View full text Add to dashboard Cite

Clusters of workstations employ flexible topologies: regular, irregular, and hierarchical topologies have been used in such systems. The flexibility poses challenges for developing efficient collective communication algorithms since the network topology can potentially have a strong impact on the communication performance. In this paper, we consider the all-to-all broadcast operation on clusters with cut-through and store-and-forward switches. We show that near-optimal all-to-all broadcast on a cluster with any topology can be achieved by only using the links in a spanning tree of the topology when the message size is sufficiently large. The result implies that increasing network connectivity beyond the minimum tree connectivity does not improve the performance of the all-to-all broadcast operation when the most efficient topology specific algorithm is used. All-to-all broadcast algorithms that achieve near-optimal performance are developed for clusters with cut-through and clusters with store-and-forward switches. We evaluate the algorithms through experiments and simulations. The empirical results confirm our theoretical finding. KEY WORDS:Collective communication; all-to-all broadcast; cluster of workstations.

show abstract

Section: Related Workmentioning

confidence: 99%

Bandwidth Efficient All-to-All Broadcast on Switched Clusters

Faraj

Patarasuk

Yuan

2007

Int J Parallel Prog

View full text Add to dashboard Cite

show abstract

“…Reflective memory network is mainly reflected by the memory board via a transmission medium for the optical fiber to other nodes and can be inserted in a variety of bus on the motherboard, such as VME, PCI and PXI [10],etc. 2.Broadcast memory Real-Time Network [11]. Each node forms a physical star topology network through the central HUB.…”

Section: Introductionmentioning

confidence: 99%

Design for the Parameters' Transfer of A Hybrid Real-Time Communication Network in TD-LTE

Cai¹,

Hu²,

Wang³

2015

2015 8th International Symposium on Computational Intelligence and Design (ISCID)

View full text Add to dashboard Cite

In TD-LTE test verification system, a proper method is needed for the parameters' transfer, which is based on software and hardware co-simulation platform. In this paper, the hybrid Real-Time communication network, which is composed of TCP/IP protocol, UDP protocol and reflective memory technology on account of hardware board, was built and a loopback test system between system simulation and link system for parameters' interface has been formed. The entire system has realized the sending and receiving of the parameters, which is fast, scalable and flexible link by either the optical fibers or LAN cable. Results show that the latency of the data could prove the key microsecond-level delay. That not only ensures the complex Real-Time of communication system, but also does not lose the parameters' information of the whole system.

show abstract

“…Many all-to-all broadcast algorithms were designed for specific network topologies that are used in parallel machines, including hypercube [8,20], mesh [15,18,21], torus [21], k-ary n-cube [20], fat tree [10], and star [14]. Work in [9] optimizes MPI collective communications, including M P I Allgather, on wide area networks.…”

Section: Related Workmentioning

confidence: 99%

Bandwidth Efficient All-to-All Broadcast on Switched Clusters

Faraj¹,

Patarasuk²,

Zhong³

2005

2005 IEEE International Conference on Cluster Computing

View full text Add to dashboard Cite

We develop an all-to-all broadcast scheme that achieves maximum bandwidth efficiency for clusters with tree topologies. Using our scheme for clusters with cut-through switches, any tree topology can support allto-all broadcast as efficiently as a single switch connecting all machines when the message size is sufficiently large. Since a tree topology can be embedded in almost any connected network, it follows that efficient all-toall broadcast can be achieved in almost all topologies, regular or irregular. To perform all-to-all broadcast efficiently on clusters with store-and-forward switches, the algorithm must minimize the communication path lengths in addition to maximizing bandwidth efficiency. This turns out to be a harder algorithmic problem. We develop schemes that give solutions to common cases for such systems. The performance of our algorithms is evaluated on Ethernet switched clusters with different topologies. The results confirm our theoretical finding. Furthermore, depending on the topology, our algorithms sometimes out-perform the topology-unaware algorithms used in MPI libraries, including MPICH and LAM/MPI, to a very large degree.

show abstract

Efficient all-to-all broadcast schemes in distributed-memory parallel computers

Cited by 3 publications

References 7 publications

Bandwidth Efficient All-to-All Broadcast on Switched Clusters

Bandwidth Efficient All-to-All Broadcast on Switched Clusters

Design for the Parameters' Transfer of A Hybrid Real-Time Communication Network in TD-LTE

Bandwidth Efficient All-to-All Broadcast on Switched Clusters

Contact Info

Product

Resources

About