Optimisation of ATLAS computing resource usage through a modern HEP Benchmark Suite via HammerCloud and Big PanDA

Natalia Szczepanek; Domenico Giordano; Ivan Glushkov; Gonzalo Menendez Borge; Alessandro Di Girolamo; Alexander Lory; Ilija Vukotic

arXiv · 2502.04853

Optimisation of ATLAS computing resource usage through a modern HEP Benchmark Suite via HammerCloud and Big PanDA

Natalia Szczepanek·Domenico Giordano·Ivan Glushkov·Gonzalo Menendez Borge·Alessandro Di Girolamo·Alexander Lory·Ilija Vukotic

View original source ↗Read source document ↗Export metadata ↓

Abstract

In April 2023, HEPScore23, the new benchmark based on HEP specific applications, was adopted by WLCG, replacing HEP-SPEC06. As part of the transition to the new benchmark, the CPU corepower published by the sites needed to be compared with the effective power observed while running ATLAS workloads. One aim was to verify the conversion rate between the scores of the old and the new benchmark. The other objective was to understand how the HEPScore performs when run on multi-core job slots, so exactly like the computing sites are being used in the production environment. Our study leverages the HammerCloud infrastructure and the PanDA Workload Management System to collect a large benchmark statistic across 136 computing sites using an enhanced HEP Benchmark Suite. It allows us to collect not only performance metrics, but, thanks to plugins, it also collects information such as machine load, memory usage and other user-defined metrics during the execution and stores it in an OpenSearch database. These extensive tests allow for an in-depth analysis of the actual, versus declared computing capabilities of these sites. The results provide valuable insights into the real-world performance of computing resources pledged to ATLAS, identifying areas for improvement while spotlighting sites that underperform or exceed expectations. Moreover, this helps to ensure efficient operational practices across sites. The collected metrics allowed us to detect and fix configuration issues and therefore improve the experienced performance.

Explore related subjects

cs.DC hep-ex

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Natalia Szczepanek, Domenico Giordano, Ivan Glushkov, Gonzalo Menendez Borge, Alessandro Di Girolamo, Alexander Lory, Ilija Vukotic. 2025-02-07. Optimisation of ATLAS computing resource usage through a modern HEP Benchmark Suite via HammerCloud and Big PanDA. https://arxiv.org/abs/2502.04853

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Co-Fabric: Breaking Host-Domain Boundaries for Unified xPU Interconnection

Large-model parameters have grown beyond the capacity of a single xPU, dispersing across multiple xPUs spanning distinct host domains, where xPU-to-xPU communication dominates overall system efficiency. Existing scale-up interconnect remains inadequate: network-based solutions built on Ethernet--such as RoCE (RDMA over Converged Ethernet)--introduce specific message-semantics and protocol-stack characteristics, and rely on fragmented per-host addressing, while conventional host-based fabrics are confined to a single host domain and lack cross-host unified addressing. This paper presents Co-Fabric, a bus-based interconnect that, unlike conventional bus designs, breaks host-domain boundaries to deliver unified xPU interconnection for scale-up superpods. Co-Fabric makes three contributions: a streamlined four-layer protocol stack achieving nanosecond-scale processing latency with native reliability; a cross-domain scaling and P2P mechanism that routes using port identifiers embedded in the packet header; and a unified address space built on shadow-device auto-enumeration. On a 64-xPU 3D-Mesh system, Co-Fabric cuts inter-node communication latency by over 50% and improves bandwidth by 2-5x over RoCE, accelerating DeepSeek R1 inference by 30%-80%. Moreover, since its streamlined four-layer protocol stack and higher data-communication efficiency reduce protocol and processing overhead relative to the Ethernet-based RoCE stack, Co-Fabric cuts the cost and power of the interconnect itself by up to 80% and 5%, respectively. These results demonstrate Co-Fabric's advantage for AI computing centers.

cs.DC↗

arXiv2026

Hydrozoan: Latency-Adaptive DAG Consensus under Mixed Byzantine and Crash Faults

DAG-based consensus protocols can achieve great throughput and the optimal three-message-delay limit for n = 3f+1 consensus. While two-delay protocols exist, they pay with reduced resilience (requiring 5f+1-style committees) or rely on fallbacks that sacrifice the DAG's high throughput. This paper introduces Hydrozoan, the first DAG protocol with a dual commit path under a hybrid fault model of f Byzantine and c crashed validators, on n = 3f+c+2p+1 validators. Leaders commit in two message delays whenever at most p validators are faulty, and in three otherwise, with no extra messages, no view changes, and multiple leaders per round. Both paths are evaluated on the same DAG, using a novel graded indirect rule to reconcile them so that every honest validator reaches the same decision. We show that under geo-distributed conditions, which path is faster is a property of geography rather than the protocol, as rounds reaching a remote region cost far more than those that do not. The (f, c, p) knobs place the fast quorum where the deployment requires it, allowing a commit in two message delays. If misconfigured, Hydrozoan can still commit in three message delays: Hydrozoan commits on whichever path fires first. We also present Optimal-Hydrozoan, a variant that tolerates one more fault on the fast path, the first construction to match the known lower bound. The safety and liveness of both protocols are machine-checked in Lean 4. Our geo-distributed evaluation shows that Hydrozoan matches Mysticeti's throughput, commits ~25% faster when the fast quorum fits fast regions, and falls back to three message delays when it does not or past p faults, where existing two-delay protocols stall.

cs.DC↗

arXiv2026

Reforge: Low-Latency Distributed GNN Serving with Selective Embedding Recomputation

Graph Neural Networks (GNNs) have been widely adopted for their ability to compute expressive node representations in graph datasets. However, serving GNNs on large graphs is challenging due to the high communication, computation, and memory overheads of constructing and executing computation graphs, which represent information flow across large neighborhoods. Existing approximation techniques in training can mitigate the overheads but, in serving, still lead to high latency and/or accuracy loss. To this end, we propose Reforge, a system that enables low-latency GNN serving for large graphs with minimal accuracy loss through two key ideas. First, Reforge employs selective recomputation of precomputed embeddings, which allows for reusing precomputed computation subgraphs while selectively recomputing a small fraction to minimize accuracy loss. Second, we develop computation graph parallelism, which reduces communication overhead by parallelizing the creation and execution of computation graphs across machines. Our evaluation with large graph datasets and GNN models shows that Reforge significantly outperforms state-of-the-art techniques.

cs.DC↗