Report of the 2025 Workshop on Next-Generation Ecosystems for Scientific Computing: Harnessing Community, Software, and AI for Cross-Disciplinary Team Science

TL;DR

The report identifies AI-enabled scientific computing as a turning point and advocates socio-technical co-design to build robust, modular software ecosystems that integrate AI with HPC, data, and human expertise. It advances three core axes—software ecosystems for AI in scientific computing, cross-disciplinary collaboration with AI for teams, and pedagogy/workforce development—while underscoring AI as a catalyst and community engagement as foundational. It proposes near-term pilots (hybrid AI/HPC infrastructure, responsible AI guidelines, and public-private partnerships) and longer-term strategies for governance, training, and global collaboration to sustain innovation. The work highlights the need for modular, trustworthy software, validated AI-generated code, scalable education, and strategic partnerships to accelerate discovery while preserving scientific rigor and equity.

Abstract

This report summarizes insights from the 2025 Workshop on Next-Generation Ecosystems for Scientific Computing: Harnessing Community, Software, and AI for Cross-Disciplinary Team Science, which convened more than 40 experts from national laboratories, academia, industry, and community organizations to chart a path toward more powerful, sustainable, and collaborative scientific software ecosystems. To address urgent challenges at the intersection of high-performance computing (HPC), AI, and scientific software, participants envisioned agile, robust ecosystems built through socio-technical co-design--the intentional integration of social and technical components as interdependent parts of a unified strategy. This approach combines advances in AI, HPC, and software with new models for cross-disciplinary collaboration, training, and workforce development. Key recommendations include building modular, trustworthy AI-enabled scientific software systems; enabling scientific teams to integrate AI systems into their workflows while preserving human creativity, trust, and scientific rigor; and creating innovative training pipelines that keep pace with rapid technological change. Pilot projects were identified as near-term catalysts, with initial priorities focused on hybrid AI/HPC infrastructure, cross-disciplinary collaboration and pedagogy, responsible AI guidelines, and prototyping of public-private partnerships. This report presents a vision of next-generation ecosystems for scientific computing where AI, software, hardware, and human expertise are interwoven to drive discovery, expand access, strengthen the workforce, and accelerate scientific progress.

Figures (6)

• Figure 1: Toward next-generation ecosystems for scientific computing: As motivated by the needs of team-based science in an AI-driven future, we must advance software, cross-disciplinary collaboration, and pedagogy (outer ring of this figure) through co-design, all while developing workforce and community. This work demands strong collaboration among researchers (in science domains, applied math, computer science, HPC, AI, etc.), software developers, stakeholders, and computing facilities (inner rings of this figure).
• Figure 2: Simulations in advanced scientific computing (including materials science, astrophysics, nuclear energy, biology, engine design, weather prediction, and batteries, as represented by the science images of this figure) require collaboration across science domains, applied mathematics, computer science, ML/AI, and more, where high-quality software is a primary means of encapsulating expertise for use by others. Due to increasing science challenges and complexity, no longer can a single small team (shown on the left-hand side) independently develop all required functionality. Instead, reusable libraries and tools (shown on the right-hand side), developed by teams whose expertise spans across various topics, provide key functionalities that serve many applications. Dashed lines indicate multiple areas of work per person (e.g., the red-colored person in this diagram contributes to application components, development tools, and ML/AI capabilities).
• Figure 3: Socio-technical co-design for next-generation scientific computing intentionally interweaves technical and social elements throughout all aspects of work, while closely coupling cycles of R&D innovation between computing technologies and driving applications. By ensuring that technical and social dimensions are addressed as tightly coupled elements of a unified, forward-looking strategy, this holistic approach accelerates transformative impact across wide-ranging application domains.
• Figure 4: Report structure.
• Figure 5: Team challenges span throughout macro-, meso-, and micro-level factors.