Simon Peter
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Simon Peter has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Thomas Anderson
University of Washington, Seattle, USA
4 shared publications - Jonggyu Park
University of Washington, Seattle, USA
3 shared publications - Aditya K. Kamath
University of Washington, Seattle, USA
2 shared publications - Baris Kasikci
University of Washington, Seattle, WA, USA
2 shared publications - Dedong Xie
University of Washington, Seattle, WA, USA
2 shared publications - Marco Canini
King Abdullah University of Science and Technology, Thuwal, Saudi Arabia
2 shared publications
Latest Publications
- Open Access
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 352-367
FM 2 : Fast Live VM Migration in CXL Memory Pools
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 352-367
Open AccessFM 2 : Fast Live VM Migration in CXL Memory Pools
- Open Access
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1360-1375
Presto: A Match-Action TCP Stack for the Terabit Era
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1360-1375
Open AccessPresto: A Match-Action TCP Stack for the Terabit Era
- Open Access
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1689-1706
Prediction-Informed Power Management for General-Purpose Compute Servers
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1689-1706
Open AccessPrediction-Informed Power Management for General-Purpose Compute Servers
- Open Access
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1365-1380
PASS: A Power Adaptive Storage Server
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1365-1380
Open AccessPASS: A Power Adaptive Storage Server
- Open Access
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 899-918
Reducing the GPU Memory Bottleneck with Lossless Compression for ML
Machine learning (ML) training and inference often process data sets far exceeding GPU memory capacity, forcing them to rely on PCIe for on-demand tensor transfers, causing critical transfer bottlenecks. Lossy compression has been proposed to relieve bottlenecks but introduces workload-dependent accuracy loss, making it complex or even prohibitive to use in existing ML deployments. We explore lossless compression as an alternative that avoids this deployment complexity. We identify where lossless compression can be integrated into ML pipelines while minimizing interference with GPU execution. Based on our findings, we introduce Invariant Bit Packing (IBP), a novel lossless compression algorithm designed to minimize data transfer time for ML. IBP identifies and eliminates invariant bits across groups of tensors, improving throughput through GPU-optimized decompression that leverages warp parallelism, low-overhead bit operations, and asynchronous PCIe transfers. We provide easy-to-use APIs, showcasing them by adding IBP support to GNN training, as well as DLRM and LLM inference frameworks. IBP achieves, on average, 74% faster GNN training, 180% faster DLRM embedding lookup, and 25% faster LLM inference.
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 899-918
Open AccessReducing the GPU Memory Bottleneck with Lossless Compression for ML
Machine learning (ML) training and inference often process data sets far exceeding GPU memory capacity, forcing them to rely on PCIe for on-demand tensor transfers, causing critical transfer bottlenecks. Lossy compression has been proposed to relieve bottlenecks but introduces workload-dependent accuracy loss, making it complex or even prohibitive to use in existing ML deployments. We explore lossless compression as an alternative that avoids this deployment complexity. We identify where lossless compression can be integrated into ML pipelines while minimizing interference with GPU execution. Based on our findings, we introduce Invariant Bit Packing (IBP), a novel lossless compression algorithm designed to minimize data transfer time for ML. IBP identifies and eliminates invariant bits across groups of tensors, improving throughput through GPU-optimized decompression that leverages warp parallelism, low-overhead bit operations, and asynchronous PCIe transfers. We provide easy-to-use APIs, showcasing them by adding IBP support to GNN training, as well as DLRM and LLM inference frameworks. IBP achieves, on average, 74% faster GNN training, 180% faster DLRM embedding lookup, and 25% faster LLM inference.
- Open Access
- 13 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 45-53
Closing the Benchmark Gap for Tiered Memory
- 13 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 45-53
Open AccessClosing the Benchmark Gap for Tiered Memory
- Open Access
- 11 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 98-105
Cloud abstractions for AI workloads
- 11 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 98-105
Open AccessCloud abstractions for AI workloads
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 897-912
POD-A ttention : Unlocking Full Prefill-Decode Overlap for Faster LLM Inference
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 897-912
Open AccessPOD-A ttention : Unlocking Full Prefill-Decode Overlap for Faster LLM Inference
- Open Access
- 8 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 47-54
Can Storage Devices be Power Adaptive?
Power is becoming a scarce resource for data centers, raising the need for power adaptive system design---the ability to dynamically change power consumption---to match available power. Storage makes up an increasing fraction of total data center power consumption. As such, it holds great potential to contribute to data center power adaptivity. To this end, we conduct a measurement study of power control mechanisms on a variety of modern data center storage devices. By changing device power states and shaping IO, we achieve a power dynamic range of up to 59.4% of the device's maximum operating power. We also study power control trade-offs, including throughput and latency. Based on our observations, we construct storage device power-throughput models and discuss the implications on power adaptive storage system design.
- 8 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 47-54
Open AccessCan Storage Devices be Power Adaptive?
Power is becoming a scarce resource for data centers, raising the need for power adaptive system design---the ability to dynamically change power consumption---to match available power. Storage makes up an increasing fraction of total data center power consumption. As such, it holds great potential to contribute to data center power adaptivity. To this end, we conduct a measurement study of power control mechanisms on a variety of modern data center storage devices. By changing device power states and shaping IO, we achieve a power dynamic range of up to 59.4% of the device's maximum operating power. We also study power control trade-offs, including throughput and latency. Based on our observations, we construct storage device power-throughput models and discuss the implications on power adaptive storage system design.