An official website of the United States government
Managing and preparing complex data for deep learning, a prevalent approach in large-scale data science can be challenging. Data transfer for model training also presents difficulties, impacting scientific fields like genomics, climate modeling, and astronomy. A large-scale solution like Google Pathways with a distributed execution environment for deep learning models exists but is proprietary. Integrating existing open-source, scalable runtime tools and data frameworks on high-performance computing (HPC) platforms is crucial to address these challenges. Our objective is to establish a smooth and unified method of combining data engineering and deep learning frameworks with diverse execution capabilities that can be deployed on various high-performance computing platforms, including cloud and supercomputers. We aim to support heterogeneous systems with accelerators, where Cylon and other data engineering and deep learning frameworks can utilize heterogeneous execution. To achieve this, we propose Radical-Cylon, a heterogeneous runtime system with a parallel and distributed data framework to execute Cylon as a task of Radical Pilot. We thoroughly explain Radical-Cylon’s design and development and the execution process of Cylon tasks using Radical Pilot. This approach enables the use of heterogeneous MPI-Communicators across multiple nodes. Radical-Cylon achieves better performance than Bare-Metal Cylon with minimal and constant overhead. Radical-Cylon achieves (4 15)% faster execution time than batch execution while performing similar join and sort operations with 35 million and 3.5 billion rows with the same resources. The approach aims to excel in both scientific and engineering research HPC systems while demonstrating robust performance on cloud infrastructures. This dual capability fosters collaboration and innovation within the open-source scientific research community.Not Available
more »
« less
This content will become publicly available on September 19, 2026
Linux and High-Performance Computing
In the 1980s, high-performance computing (HPC) became another tool for research in the open (non-defense) science and engineering research communities. However, HPC came with a high price tag; the first Cray-2 machines, released in 1985, cost between $12 million and $17 million, according to the Computer History Museum, and were largely available only at government research labs or through national supercomputing centers. In the 1990s, with demand for HPC increasing due to vast datasets, more complex modeling, and the growing computational needs of scientific applications, researchers began experimenting with building HPC machines from clusters of servers running the Linux operating system. By the late 1990s, two approaches to Linux-based parallel computing had emerged: the personal computer cluster methodology that became known as Beowulf and the Roadrunner architecture aimed at a more cost-effective supercomputer. While Beowulf attracted attention because of its low cost and thereby greater accessibility, Roadrunner took a different approach. While still affordable compared to vector processors and other commercially available supercomputers, Roadrunner integrated its commodity components with specialized networking technology. Furthermore, these systems initially served different purposes. While Beowulf focused on providing affordable parallel workstations for individual researchers at NASA, Roadrunner set out to provide a multiuser system that could compete with the commercial supercomputers that dominated the market at the time. This paper analyzes the technical decisions, performance implications, and long-term influence of both approaches. Through this analysis, we can start to judge the impact of both Roadrunner and Beowulf on the development of Linux-based supercomputers.
more »
« less
- PAR ID:
- 10655202
- Publisher / Repository:
- Techrxiv
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
The next generation of supercomputing resources is expected to greatly expand the scope of HPC environments, both in terms of more diverse workloads and user bases, as well as the integration of edge computing infrastructures. This will likely require new mechanisms and approaches at the Operating System level to support these broader classes of workloads along with their different security requirements. We claim that a key mechanism needed for these workloads is the ability to securely compartmentalize the system software executing on a given node. In this paper, we present initial efforts in exploring the integration of secure and trusted computing capabilities into an HPC system software stack. As part of this work we have ported the Kitten Lightweight Kernel (LWK) to the ARM64 architecture and integrated it with the Hafnium hypervisor, a reference implementation of a secure partition manager (SPM) that provides security isolation for virtual machines. By integrating Kitten with Hafnium, we are able to replace the commodity oriented Linux based resource management infrastructure and reduce the overheads introduced by using a full weight kernel (FWK) as the node-level resource scheduler. While our results are very preliminary, we are able to demonstrate measurable performance improvements on small scale ARM based SOC platforms.more » « less
-
Between May 25, 2023 and June 21, 2023, we hosted the inaugural four-week High-Performance Computing Summer Institute at Jackson State University. This endeavor was made possible through the support of a three-year NSF CISE-MSI grant. The primary objective of this Summer Institute revolved around the engagement, education, and empowerment of minority and underrepresented students in the realm of High-Performance Computing (HPC) within the field of engineering. Nine undergraduate students with diverse background were recruited to participate in this program. Throughout the program, we immersed these students in a comprehensive curriculum that covered various critical facets of HPC. This curriculum encompassed hands-on instruction in Linux operating system command-line operations, C programming within the Linux environment, fundamental HPC concepts, parallel computing utilizing the Message Passing Interface (MPI) library, and GPU computing through OpenCL. Additionally, we delved into foundational aspects of fluid mechanics, geometric modeling, mesh generation, flow simulation via our in-house flow solvers, and the visualization of solutions. At the end of the program, every participant was tasked with delivering an oral presentation and submitting a written report encapsulating their acquired knowledge and experiences during the program. We are excited to share a detailed overview of our program's implementation with our audience. This includes insights into our utilization of ChatGPT to enhance C programming learning and our suggestion of the NSF ACCESS resources to gain access to HPC systems. We are proud to announce that the program has achieved remarkable success, as evidenced by the positive feedback we received from the participants.more » « less
-
High performance computing (HPC) system runs compute-intensive parallel applications requiring large number of nodes. An HPC system consists of heterogeneous computer architecture nodes, including CPUs, GPUs, field programmable gate arrays (FPGAs), etc. Power capping is a method to improve parallel application performance subject to variable power constraints. In this paper, we propose a parallel application power and performance prediction simulator. We present prediction model to predict application power and performance for unknown power-capping values considering heterogeneous computing architecture. We develop a job scheduling simulator based on parallel discrete-event simulation engine. The simulator includes a power and performance prediction model, as well as a resource allocation model. Based on real-life measurements and trace data, we show the applicability of our proposed prediction model and simulator.more » « less
-
In this article, we present a four-layer distributed simulation system and its adaptation to the Material Point Method (MPM). The system is built upon a performance portableC++programming model targeting major High-Performance-Computing (HPC) platforms. A key ingredient of our system is a hierarchical block-tile-cell sparse grid data structure that is distributable to an arbitrary number of Message Passing Interface (MPI) ranks. We additionally propose strategies for efficient dynamic load balance optimization to maximize the efficiency of MPI tasks. Our simulation pipeline can easily switch among backend programming models, including OpenMP and CUDA, and can be effortlessly dispatched onto supercomputers and the cloud. Finally, we construct benchmark experiments and ablation studies on supercomputers and consumer workstations in a local network to evaluate the scalability and load balancing criteria. We demonstrate massively parallel, highly scalable, and gigascale resolution MPM simulations of up to 1.01 billion particles for less than 323.25 seconds per frame with 8 OpenSSH-connected workstations.more » « less
