High Performance MPI on Infiniband Cluster

Overview

The MVAPICH2 software, supporting MPI 3.0 standard, delivers the best performance, scalability, and fault tolerance for high-end computing systems and servers using InfiniBand, Omni-Path, Ethernet/iWARP, and RoCE networking technologies. This software is being used by more than 3,125 organizations world-wide in 89 countries to extract the potential of these emerging networking technologies for modern systems.

Description

MVAPICH2 provides many features including MPI-3 standard compliance, single copy intra-node communication using Linux supported CMA (Cross Memory Attach), Checkpoint/Restart using LLNL's Scalable Checkpoint/Restart Library (SCR), high-performance and scalable InfiniBand hardware multicast-based collectives, enhanced shared-memory-aware and intra-node Zero-Copy collectives (using LiMIC), high-performance communication support for NVIDIA GPU with IPC, collective and non-contiguous datatype support, integrated hybrid UD-RC/XRC design, support for UD only mode, nemesis-based interface, shared memory interface, scalable and robust daemon-less job startup (mpirun-rsh), flexible process manager support (mpirun-rsh and hydra.mpiexec), full autoconf-based configuration, portable hardware locality (hwloc) with flexible CPU granularity policies (core, socket and numanode) and binding policies (bunch and scatter) with SMT support, flexible rail binding with processes for multirail configurations, message coalescing, dynamic process migration, fast process-level fault-tolerance with checkpoint-restart, fast job-pause-migration-resume framework for pro-active fault-tolerance, suspend/resume, network-level fault-tolerance with Automatic Path Migration (APM), RDMA CM support, iWARP support, optimized collectives, on-demand connection management, multi-pathing, RDMA Read-based and RDMA-write-based designs, polling and blocking-based communication progress, multi-core optimized and scalable shared memory support, LiMIC2-based kernel-level shared memory support for both two-sided and one-sided operations, shared memory backed Windows for one-Sided communication, HugePage support, and memory hook with ptmalloc2 library support. The ADI-3-level design of MVAPICH2 2.1rc1 supports many features including: MPI-2 functionalities (one-sided, dynamic process management, collectives and datatype), multi-threading and all MPI-1 functionalities. It also supports a wide range of platforms, architectures, OS, compilers, InfiniBand adapters (Mellanox and QLogic), iWARP adapters (including the new Chelsio T4 adapter) and RoCE adapters.

Journals (11)

1 K. Suresh, K. Khorassani, C. Chen, B. Ramesh, M. Abduljabbar, A. Shafi, H. Subramoni, and DK Panda, Network Assisted Non-Contiguous Transfers for GPU-Aware MPI Libraries, IEEE Micro, Jan 2023.
2 S. Ramesh, A. Mahéo, S. Shende, A. Malony, H. Subramoni, A. Ruhela, and DK Panda, MPI performance engineering with the MPI tool interface: The integration of MVAPICH and TAU, ISSN 0167-8191, Volume 77, Sep 2018.
3 S. Sur, S. Potluri, K. Kandalla, H. Subramoni, K. Tomko, and DK Panda, Co-Designing MPI Library and Applications for InfiniBand Clusters IEEE Computer, Nov 2011.
4 A. Vishnu, M. Koop, A. Moody, A. Mamidala, S. Narravula, and DK Panda, Topology Agnostic Hot-Spot Avoidance with InfiniBand Concurrency and Computation: Practice and Experience, Special Issue of Best Papers from CCGrid '07, Jan 2008.
5 H. Wang, S. Potluri, D. Bureddy, and DK Panda, GPU-Aware MPI on RDMA-Enabled Cluster: Design, Implementation and Evaluation, IEEE Transactions on Parallel & Distributed Systems, Vol. 25, No. 10, pp. 2595-2605, Oct 2014.
6 C. Chu, X. Lu, Ammar Awan, H. Subramoni, Bracy Elton, and DK Panda, Exploiting Hardware Multicast and GPUDirect RDMA for Efficient Broadcast, IEEE Transactions on Parallel and Distributed Systems (TPDS), vol. 30, no. 3, pp. 575-588, 1 March 2019,
7 Ammar Awan, K. Vadambacheri Manian, C. Chu, H. Subramoni, and DK Panda, Optimized Large-Message Broadcast for Deep Learning Workloads: MPI, MPI+NCCL, or NCCL2?, Volume 85, July 2019, Pages 141-152, https://doi.org/10.1016/j.parco.2019.03.005,
8 S. Chakraborty, Ignacio Laguna, Murali Emani, Kathryn Mohror, DK Panda, Martin Schulz, and H. Subramoni, EReinit: Scalable and Efficient Fault Tolerance for Bulk-Synchronous MPI Applications, Concurrency and Computation: Practice and Experience, 14 August 2018, https://doi.org/10.1002/cpe.4863,
9 A. Ruhela, H. Subramoni, S. Chakraborty, M. Bayatpour, P. Kousha, and DK Panda, Effcient Design for MPI Asynchronous Progress without Dedicated Resources, Parallel Computing - Systems & Applications, Volume 85, July 2019, Pages 13-26, https://doi.org/10.1016/j.parco.2019.03.003,
10 Ammar Awan, A. Jain, C. Chu, H. Subramoni, and DK Panda, Communication Profiling and Characterization of Deep Learning Workloads on Clusters with High-Performance Interconnects, IEEE Micro, vol. 40, no. 1, pp. 35-43, 1 Jan.-Feb. 2020.,
11 J. Hashmi, C. Chu, S. Chakraborty, M. Bayatpour, H. Subramoni, and DK Panda, FALCON-X: Zero-copy MPI Derived Datatype Processing on Modern CPU and GPU Architectures, Journal of Parallel and Distributed Computing (JPDC), Volume 144, October 2020, Pages 1-13, doi.org/10.1016/j.jpdc.2020.05.008,

Conferences & Workshops (293)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293

Ph.D. Disserations (1)

1 S. Potluri, Enabling Efficient Use of MPI and PGAS Programming Models on Heterogeneous Clusters with High Performance Interconnects, May 2014

M.S. Thesis (3)

1 V. Dhanraj, Enhancement of LIMIC-Based Collectives for Multi-core Clusters, Aug 2012
2 A. Singh, Optimizing All-to-all and Allgather Communications on GPGPU Clusters, Apr 2012
3 K. Gopalakrishnan, Enhancing Fault Tolerance in MPI for Modern InfiniBand Clusters, Aug 2009