As experimental methods continue to evolve, generating increasingly complex and high-resolution datasets, machine learning (ML) is becoming an essential tool across numerous scientific disciplines. This conference will explore emerging ML methods and their applications in scientific discovery, focusing on processing technologies and strategies to accelerate deep learning and inference.
Fee Waivers: Thanks to generous support from our sponsors, the NSF HDR Institute AI-Accelerated Algorithms for Data-Driven Discovery (A3D3) and Exploring Neural Network Processors for AI in Science and Engineering (Voyager), we are offering in-person registration fee waivers (subject to funding availability) for graduate students on a first come first serve basis. If funds are still available, we will also open up fee waivers to undergraduate students and early-career researchers (<7 years since PhD) with financial need. Please indicate in the registration form if you are seeking a registration fee waiver and do not pay the fee.
Important Deadlines

• Abstract Submission: June 1, 2026
• (Optional) Extended Abstract Submission: June 15, 2026
• Early Bird Registration: July 15, 2026
  
  Topics include, but are not limited to:
  - Machine Learning Algorithm Design & Optimization
  - Accelerated Inference & Real-Time Processing
  - Scientific Applications of Fast ML
  - Scalable & Distributed ML Systems
  - Advanced Hardware & Computing Architectures 

We welcome abstracts for:

• Presentations and/or Posters
• • New for 2026: Submitters will have the option to submit a 4-page extended abstract (paper) to OpenReview (details to follow)
• Tutorials
• Topical (birds-of-a-feather) sessions

More information and registration details will follow. We look forward to welcoming you to San Diego this September!

• Venue: Online (Part of 2026 Science Gateways Conference)
• When: Aug 31, 2026, 1 pm-2:30 pm ET
• Duration: 90 minutes
• Registration: Free registration is required to attend the tutorial.
• Skill level: Beginner to intermediate. No programming experience required.
• Requirements: Web browser. Tutorial accounts and preconfigured compute resources will be provided.

About This Tutorial
The NSF ACCESS and NAIRR ecosystems provide researchers and educators with access to advanced computing, data, and AI resources, yet many users still face practical barriers to using command-line workflows common on HPC and AI resources. This tutorial introduces iDLab, an NSF-funded effort to advance federated interactive computing and data environments, and OneSciencePlace, the composable platform that enables the creation of science gateways, HPC/AI portals, research applications, and data publication environments via a low-code/no-code interface.
Participants will learn how OneSciencePlace enables applications to be deployed and accessed via a web browser by individual researchers, classrooms, research groups, campus programs, and broader consortia. The tutorial will focus on hands-on use and deployment of applications on selected ACCESS and NAIRR resources, aligning with iDLab's goal of making advanced computing and data resources easier to use across institutions.
Examples will include AI inference and machine learning classification via interactive web applications such as Jupyter, interactive VNC applications such as Linux Desktop, and batch command-line applications on HPC systems. Participants will also review more complex application interfaces and create a simple app with a custom web form for parameters, inputs, and execution.
This tutorial builds on two well-attended OneSciencePlace tutorials delivered at Gateways 2024 and Gateways 2025. It extends them by placing greater emphasis on deploying applications on ACCESS and NAIRR resources, introducing newer OneSciencePlace capabilities, and situating the work within the broader iDLab effort.

---

Learning Objectives
By the end of the tutorial, attendees will be able to:

1. Run web, VNC, and batch applications through OneSciencePlace
2. Understand how iDLab and OneSciencePlace can connect user-facing applications to ACCESS and NAIRR resources
3. Create a simple application with a custom no-code interface
4. Understand how job parameters and execution provenance support relaunching, reproducibility, and iteration
5. Recognize how applications can be organized as FAIR, discoverable, reusable, and citable gateway objects
6. Evaluate how iDLab and OneSciencePlace can support use cases such as a single-user portal, classroom environment, research group gateway, campus service, or multi-institution consortium

This hands-on workshop provides a practical introduction to command-line Linux and high-performance computing (HPC), tailored for students and researchers new to computational environments. Participants will learn the fundamentals of using the Linux command line for navigating directories, manipulating files, and executing basic shell commands. This session also covers essential skills for working on HPC systems, including secure file transfer with tools like scp and rsync, connecting to remote clusters via SSH, and basic file manipulation using vi. By the end of the workshop, attendees will be equipped to effectively interact with ARCC2, UC’s research computing cluster, and have the foundation necessary to begin implementing complex research workflows.

This workshop introduces participants to SLURM, a widely used job scheduling system for high-performance computing (HPC) environments, and is designed as a follow-up to the Intro to Linux/Intro to HPC session. Familiarity with Linux command line is expected. Attendees will learn how to submit, monitor, and manage computational jobs on UC’s supercomputing cluster ARCC2 using core SLURM commands such as sbatch, squeue, scancel, and salloc. The workshop will cover key concepts including partitions, job scripts, resource requests, and job arrays, providing practical examples to help users optimize their computing workflows. By the end of the workshop, attendees will have the foundational tools needed to work effectively in SLURM-managed environments.