FatPaths: Routing in Supercomputers and Data Centers when Shortest Paths Fall Short

Besta, Maciej; Schneider, Marcel; Konieczny, Marek; Cynk, Karolina; Henriksson, Erik; Girolamo, Salvatore Di; Singla, Ankit; Hoefler, Torsten

doi:10.1109/sc41405.2020.00031

Cited by 17 publications

(22 citation statements)

References 130 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, we were still able to extensively simulate all the considered topologies of sizes around 10,000 endpoints, with the full TCP/IP stack. Extensive results are elsewhere [20].…”

Section: Performance Analysismentioning

confidence: 98%

See 1 more Smart Citation

Towards Million-Server Network Simulations on Just a Laptop

Besta¹,

Schneider²,

Girolamo³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The growing size of data center and HPC networks pose unprecedented requirements on the scalability of simulation infrastructure. The ability to simulate such large-scale interconnects on a simple PC would facilitate research efforts. Unfortunately, as we first show in this work, existing shared-memory packet-level simulators do not scale to the sizes of the largest networks considered today. We then illustrate a feasibility analysis and a set of enhancements that enable a simple packet-level htsim simulator to scale to the unprecedented simulation sizes on a single PC. Our code is available online and can be used to design novel schemes in the coming era of omnipresent data centers and HPC clusters.

show abstract

“…However, we were still able to extensively simulate all the considered topologies of sizes around 10,000 endpoints, with the full TCP/IP stack. Extensive results are elsewhere [20].…”

Section: Performance Analysismentioning

confidence: 98%

“…We use three-stage diameter-4 FTs; fewer stages reduce scalability while more stages lead to high diameters. Some works exist into the routing of such networks, with simulation data available [8,11,20].…”

Section: Networkmentioning

confidence: 99%

Towards Million-Server Network Simulations on Just a Laptop

Besta¹,

Schneider²,

Girolamo³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…To keep the coloring consistent with respect to already colored partitions, we maintain bitmaps B v that indicate colors already taken by v's neighbors in already colored partitions: If v cannot use a color c, the c-th bit in B v is set to 1. These bitmaps are updated before each call of SIM-COL (Lines [16][17][18].…”

Section: B Graph Coloring By Silent Conflict Resolution (Dec-adg)mentioning

confidence: 99%

High-Performance Parallel Graph Coloring with Strong Guarantees on Work, Depth, and Quality

Besta

Carigiet

Janda

et al. 2020

SC20: International Conference for High Performance Computing, Networking, Storage and Analysis

Self Cite

View full text Add to dashboard Cite

We develop the first parallel graph coloring heuristics with strong theoretical guarantees on work and depth and coloring quality. The key idea is to design a relaxation of the vertex degeneracy order, a well-known graph theory concept, and to color vertices in the order dictated by this relaxation. This introduces a tunable amount of parallelism into the degeneracy ordering that is otherwise hard to parallelize. This simple idea enables significant benefits in several key aspects of graph coloring. For example, one of our algorithms ensures polylogarithmic depth and a bound on the number of used colors that is superior to all other parallelizable schemes, while maintaining workefficiency. In addition to provable guarantees, the developed algorithms have competitive run-times for several real-world graphs, while almost always providing superior coloring quality. Our degeneracy ordering relaxation is of separate interest for algorithms outside the context of coloring.

show abstract

“…In Dragonfly networks (including SLINGSHOT), any pair of nodes is connected by multiple minimal and non-minimal paths [12], [18]. For example, by considering the topology in Figure 3, the minimal path connecting N0 to N496 includes the switches S0 and S31.…”

Section: Routingmentioning

confidence: 99%

“…For example, by considering the topology in Figure 3, the minimal path connecting N0 to N496 includes the switches S0 and S31. In smaller networks, due to links redundancy,multiple minimal paths are connecting any pair of nodes [18]. On the other hand, a possible non-minimal path involves an intermediate switch that is directly connected to both S0 and S31.…”

Section: Routingmentioning

confidence: 99%

An In-Depth Analysis of the Slingshot Interconnect

Sensi¹,

Girolamo²,

McMahon³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

The interconnect is one of the most critical components in large scale computing systems, and its impact on the performance of applications is going to increase with the system size. In this paper, we will describe SLINGSHOT, an interconnection network for large scale computing systems. SLINGSHOT is based on high-radix switches, which allow building exascale and hyperscale datacenters networks with at most three switch-to-switch hops. Moreover, SLINGSHOT provides efficient adaptive routing and congestion control algorithms, and highly tunable traffic classes. SLINGSHOT uses an optimized Ethernet protocol, which allows it to be interoperable with standard Ethernet devices while providing high performance to HPC applications. We analyze the extent to which SLINGSHOT provides these features, evaluating it on microbenchmarks and on several applications from the datacenter and AI worlds, as well as on HPC applications. We find that applications running on SLINGSHOT are less affected by congestion compared to previous generation networks.

show abstract

FatPaths: Routing in Supercomputers and Data Centers when Shortest Paths Fall Short

Cited by 17 publications

References 130 publications

Towards Million-Server Network Simulations on Just a Laptop

Towards Million-Server Network Simulations on Just a Laptop

High-Performance Parallel Graph Coloring with Strong Guarantees on Work, Depth, and Quality

An In-Depth Analysis of the Slingshot Interconnect

Contact Info

Product

Resources

About