OpenFAM: A Library for Programming Disaggregated Memory

Singhal, Sharad; Crasta, Clarete R.; Abdulla, Mashood; Barmawer, Faizan; Emberson, Dave; Ahobala, Ramya; Bhat, Gautham; Rajak, Rishi kesh K.; Soumya, P. N.

doi:10.1007/978-3-031-04888-3_2

Cited by 1 publication

(10 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus the overhead in the software stack is about 1 μs, consistent with earlier results 13 obtained on the InfiniBand cluster.…”

Section: Initial Results On Slingshotsupporting

confidence: 90%

“…Detailed measurements of API-level performance of OpenFAM were provided earlier. 13 We briefly review those results, and present additional results obtained as a result of optimizations and additional functionality introduced into the implementation since then. The reader is referred to the earlier implementation 13 of the API including scatter-gather, FAM atomics, and metadata operations.…”

Section: Implementation Performancementioning

confidence: 99%

“…13 We briefly review those results, and present additional results obtained as a result of optimizations and additional functionality introduced into the implementation since then. The reader is referred to the earlier implementation 13 of the API including scatter-gather, FAM atomics, and metadata operations. As mentioned earlier, 13 the implementation can be compiled to enable profiling.…”

Section: Implementation Performancementioning

confidence: 99%

“…We describe the architecture of a reference implementation 12 for OpenFAM, and provide a summary of its performance. 13 Next, we focus on the API extensions we are making. These include support for thread safety, I/O contexts, data item interleaving, support for archival storage, and memory side operations.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper, we briefly review the OpenFAM API specification, and summarize extensions we are making to the API to add functionality and improve performance. We describe the architecture of a reference implementation 12 for OpenFAM, and provide a summary of its performance 13 . Next, we focus on the API extensions we are making.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

OpenFAM: Programming disaggregated memory

Singhal,

Crasta,

Abdulla K

et al. 2023

Concurrency and Computation

View full text Add to dashboard Cite

High performance computing (HPC) clusters are increasingly handling workloads where working data sets cannot be easily partitioned or are too large to fit into local node memory. In order to enable HPC workloads to access memory external to the node, HPE has defined a programming API (OpenFAM) for developing applications that use large‐scale disaggregated memory. In this paper we describe an open‐source reference implementation of OpenFAM that can be used on scale‐up machines, traditional HPC clusters, as well as emerging disaggregated memory architectures. We demonstrate the efficiency of the implementation using micro‐benchmarks on InfiniBand and Slingshot‐based clusters.

show abstract

“…Thus the overhead in the software stack is about 1 μs, consistent with earlier results 13 obtained on the InfiniBand cluster.…”

Section: Initial Results On Slingshotsupporting

confidence: 90%

Section: Implementation Performancementioning

confidence: 99%