Michael E. Papka

We present the Federated Inference Resource Scheduling Toolkit (FIRST), a framework enabling Inference-as-a-Service across distributed High-Performance Computing (HPC) clusters. FIRST provides cloud-like access to diverse AI models, like Large Language Models (LLMs), on existing HPC infrastructure. Leveraging Globus Auth and Globus Compute, the system allows researchers to run parallel inference workloads via an OpenAI-compliant API on private, secure environments. This cluster-agnostic API allows requests to be distributed across federated clusters, targeting numerous hosted models. FIRST supports multiple inference backends (e.g., vLLM), auto-scales resources, maintains "hot" nodes for low-latency execution, and offers both high-throughput batch and interactive modes. The framework addresses the growing demand for private, secure, and scalable AI inference in scientific workflows, allowing researchers to generate billions of tokens daily on-premises without relying on commercial cloud infrastructure.

High-performance computing (HPC) systems are essential for scientific discovery and engineering innovation. However, their growing power demands pose significant challenges, particularly as systems scale to the exascale level. Prior uncore frequency tuning studies have primarily focused on conventional HPC workloads running on CPU-only systems. As HPC advances toward heterogeneous computing, integrating diverse GPU workloads on heterogeneous CPU-GPU systems, it becomes imperative to revisit and enhance uncore scaling. Our investigation reveals that uncore frequency scales down only when CPU power approaches its thermal design power (TDP), which is rare in GPU-dominant applications. As a result, modern computing systems experience unnecessary power waste. In this study, we present MAGUS, a user-transparent uncore frequency scaling runtime for heterogeneous computing. MAGUS dynamically adjusts uncore frequencies according to distinct application execution phases, effectively minimizing power waste caused by consistently using maximum uncore frequencies. Our design incorporates several key techniques, including real-time monitoring and prediction of memory accesses, intelligent handling of frequent phase transitions, and leveraging vendor-provided power management features. We evaluate MAGUS with various GPU benchmarks and applications on multiple heterogeneous systems with different CPU and GPU architectures. Experimental results demonstrate that MAGUS achieves up to 27% energy savings compared to the default settings, while maintaining a performance loss of less than 5% and an overhead of under 1%.

September 2023 marks the 50th anniversary of the Electronic Visualization Laboratory (EVL) at University of Illinois Chicago (UIC). EVL’s introduction of the CAVE Automatic Virtual Environment in 1992, the first widely replicated, projection-based, walk-in, virtual reality (VR) system in the world, put EVL at the forefront of collaborative, immersive data exploration and analytics. However, the journey did not begin then. Since its founding in 1973, EVL has been developing tools and techniques for real-time, interactive visualizations—pillars of VR. But EVL’s culture is also relevant to its successes, as it has always been an interdisciplinary lab that fosters teamwork, where each person’s expertise contributes to the development of the necessary tools, hardware, system software, applications, and human interface models to solve problems. Over the years, as multidisciplinary collaborations evolved and advanced scientific instruments and data resources were distributed globally, the need to access and share data and visualizations while working with colleagues, local and remote, synchronous and asynchronous, also became important fields of study. This paper is a retrospective of EVL’s past 50 years that surveys the many networked, immersive, collaborative visualization and VR systems and applications it developed and deployed, as well as lessons learned and future plans.