English
Related papers

Related papers: Towards CXL Resilience to CPU Failures

200 papers

Ensuring resilience in distributed systems has become an acute concern. In today's environment, it is crucial to develop light-weight mechanisms that recover a distributed system from faults quickly and with only a small impact on the…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-02-10 Antonis Psistakis , Burak Ocalan , Fabien Chaix , Ramnatthan Alagappan , Josep Torrellas

The growing prevalence of data-intensive workloads, such as artificial intelligence (AI), machine learning (ML), high-performance computing (HPC), in-memory databases, and real-time analytics, has exposed limitations in conventional memory…

The Compute Express Link (CXL) interconnect enables compute "pods" that pool memory across servers to reduce cost and improve efficiency. These pods also facilitate pairwise communication whose needs conflict with pooling. Importantly,…

Hardware Architecture · Computer Science 2026-04-06 Yuhong Zhong , Fiodar Kazhamiaka , Pantea Zardoshti , Shuwei Teng , Rodrigo Fonseca , Mark D. Hill , Daniel S. Berger

Memory disaggregation via CXL enables multi-host resource sharing. However, existing CXL sharing mechanisms enforce coarse-grained, host-level permissions only, leaving isolation to the operating system. Today, virtual memory enables…

Hardware Architecture · Computer Science 2026-05-29 Kaustav Goswami , Sean Peisert , Venkatesh Akella , Jason Lowe-Power

The substantial memory requirements of Large Language Models (LLMs), particularly for long-context fine-tuning, have renewed interest in CPU offloading to augment limited GPU memory. However, as context lengths grow, relying on CPU memory…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-10-31 Yong-Cheng Liaw , Shuo-Han Chen

Modern ML training and inference now span tens to tens of thousands of GPUs, where network faults can waste 10--15\% of GPU hours due to slow recovery. Common network errors and link fluctuations trigger timeouts that often terminate entire…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-01-01 Wei Wang , Nengneng Yu , Sixian Xiong , Zaoxing Liu

CXL-based Computational Memory (CCM) enables near-memory processing within expanded remote memory, presenting opportunities to address data movement costs associated with disaggregated memory systems and to accelerate overall performance.…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-12-05 Suyeon Lee , Kangkyu Park , Kwangsik Shin , Ada Gavrilovska

Large-scale AI training and inference require hundreds of gigabytes to terabytes of DRAM with high peak to average utilization ratios, resulting in overprovisioning. In cloud computing, DRAM constitutes a significant share of the cost. Yet,…

Hardware Architecture · Computer Science 2026-05-28 Kaustav Goswami , Maryam Babaie , Hoa Nguyen , Venkatesh Akella , Jason Lowe-Power

Large-scale LLM training requires collective communication libraries to exchange data among distributed GPUs. As a company dedicated to building and operating large-scale GPU training clusters, we encounter several challenges when using…

Concurrency has been a subject of study for more than 50 years. Still, many developers struggle to adapt their sequential code to be accessed concurrently. This need has pushed for generic solutions and specific concurrent data structures.…

Distributed, Parallel, and Cluster Computing · Computer Science 2019-11-06 Andreia Correia , Pedro Ramalhete , Pascal Felber

The observed and expected continued growth in the number of nodes in large-scale parallel computers gives rise to two major challenges: global communication operations are becoming major bottlenecks due to their limited scalability, and the…

Distributed, Parallel, and Cluster Computing · Computer Science 2020-02-03 Markus Levonyak , Christina Pacher , Wilfried N. Gansterer

Machine Learning (ML) is making a strong resurgence in tune with the massive generation of unstructured data which in turn requires massive computational resources. Due to the inherently compute- and power-intensive structure of Neural…

Machine Learning · Computer Science 2018-06-27 Behzad Salami , Osman Unsal , Adrian Cristal

Deduplication has been largely employed in distributed storage systems to improve space efficiency. Traditional deduplication research ignores the design specifications of shared-nothing distributed storage systems such as no central…

Distributed, Parallel, and Cluster Computing · Computer Science 2018-03-22 Awais Khan , Chang-Gyu Lee , Prince Hamandawana , Sungyong Park , Youngjae Kim

This paper investigates an edge computing system where requests are processed by a set of replicated edge servers. We investigate a class of applications where similar queries produce identical results. To reduce processing overhead on the…

Emerging Technologies · Computer Science 2024-05-29 Adrian-Cristian Nicolaescu , Spyridon Mastorakis , Md Washik Al Azad , David Griffin , Miguel Rio

Current HPC systems provide memory resources that are statically configured and tightly coupled with compute nodes. However, workloads on HPC systems are evolving. Diverse workloads lead to a need for configurable memory resources to…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-03-23 Jacob Wahlgren , Maya Gokhale , Ivy B. Peng

In this thesis, we introduce replay clocks (RepCl), a novel clock infrastructure that allows us to do offline analyses of distributed computations. The replay clock structure provides a methodology to replay a computation as it happened,…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-07-02 Ishaan Lagwankar

Recently, we saw the emergence of consensus-based database systems that promise resilience against failures, strong data provenance, and federated data management. Typically, these fully-replicated systems are operated on top of a…

Databases · Computer Science 2020-11-04 Suyash Gupta , Jelle Hellings , Mohammad Sadoghi

Fault tolerance overhead of high performance computing (HPC) applications is becoming critical to the efficient utilization of HPC systems at large scale. HPC applications typically tolerate fail-stop failures by checkpointing. Another…

Distributed, Parallel, and Cluster Computing · Computer Science 2011-06-22 Erlin Yao , Mingyu Chen , Rui Wang , Wenli Zhang , Guangming Tan

Data availability is critical in distributed storage systems, especially when node failures are prevalent in real life. A key requirement is to minimize the amount of data transferred among nodes when recovering the lost or unavailable data…

Distributed, Parallel, and Cluster Computing · Computer Science 2013-06-07 Runhui Li , Jian Lin , Patrick P. C. Lee

As we have entered Exascale computing, the faults in high-performance systems are expected to increase considerably. To compensate for a higher failure rate, the standard checkpoint/restart technique would need to create checkpoints at a…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-10-26 Sarthak Joshi , Sathish Vadhiyar