Toward Agentic HPC: Serving and Evaluating LLM-Powered Agents for Scientific Applications On Leadership-Class Platforms
Explore the source record for details and available documents.
Yin, Junqi [ORNL] (ORCID:0000000338435520)↗
Engineering topics
Shankar, Mallikarjun (Arjun) [ORNL] (ORCID:0000000152897460)
Publications and source records attributed to Shankar, Mallikarjun (Arjun) [ORNL] (ORCID:0000000152897460).
Data Readiness for Scientific AI at Scale
This paper examines how Data Readiness for AI (DRAI) principles apply to leadership-scale scientific datasets used to train foundation models. We analyze archetypal workflows across four representative domains—climate, nuclear fusion, bio/health, and materials—to identify common preprocessing patterns and domain-specific constraints. We introduce a two-dimensional readiness framework that combines canonical preprocessing patterns with a five-level operational readiness scale, both tailored to high-performance computing (HPC) environments. This framework helps outline key challenges in transforming large-scale scientific data into formats suitable for scalable AI training. Together, these dimensions form a conceptual maturity matrix that characterizes scientific data readiness and guides infrastructure development toward standardized, cross-domain support for scalable and reproducible AI for science.
Brewer, Wes [ORNL] (ORCID:0000000236393956)↗
Enabling Seamless Transitions from Experimental to Production HPC for Interactive Workflows
The evolving landscape of scientific computing requires seamless transitions from experimental to production HPC environments for interactive workflows. This paper presents a structured transition pathway developed at OLCF that bridges the gap between development testbeds and production systems. We address both technological and policy challenges, introducing frameworks for data streaming architectures, secure service interfaces, and adaptive resource scheduling for time-sensitive workloads and improved HPC interactivity. Our approach transforms traditional batch-oriented HPC into a more dynamic ecosystem capable of supporting modern scientific workflows that require near real-time data analysis, experimental steering, and cross-facility integration.
Etz, Brian [ORNL] (ORCID:0000000208554863)↗
RingX: Scalable Parallel Attention for Long-Context Learning on HPC
The attention mechanism has become foundational for remarkable AI breakthroughs since the introduction of the Transformer, driving the demand for increasingly longer context to power frontier models such as large-scale reasoning language models and high-resolution image/video generators. However, its quadratic computational and memory complexities present substantial challenges. Current state-of-the-art parallel attention methods, such as ring attention, are widely adopted for long-context training but utilize a point-to-point communication strategy that fails to fully exploit the capabilities of modern HPC network architectures. In this work, we propose ringX, a scalable family of parallel attention methods optimized explicitly for HPC systems. By enhancing workload partitioning, refining communication patterns, and improving load balancing, ringX achieves up to 3.4 × speedup compared to conventional ring attention on the Frontier supercomputer. Optimized for both bi-directional and causal attention mechanisms, ringX demonstrates its effectiveness through training benchmarks of a Vision Transformer (ViT) on a climate dataset and a Generative Pre-Trained Transformer (GPT) model, Llama3 8B. Our method attains an end-to-end training speedup of approximately 1.5 × in both scenarios. To our knowledge, the achieved 38% model FLOPs utilization (MFU) for training Llama3 8B with a 1M-token sequence length on 4,096 GPUs represents one of the highest training efficiencies reported for long-context learning on HPC systems. Our code implementation is available at https://github.com/jqyin/ringX-attention.
Yin, Junqi [ORNL] (ORCID:0000000338435520)↗