# Important Dates

Paper submission deadline: August 15, 2026 (AoE)
Author notification: September 4, 2026
Camera-ready deadline, with AD/AE: September 25, 2026 (AoE)

# About the Workshop

The ECHO workshop brings together researchers and practitioners across edge systems, cloud platforms, and high-performance computing (HPC) to formulate and advance a shared research agenda. In this workshop, we explore how HPC, edge, and cloud ecosystems can mutually advance physically dispersed, field-to-cloud, geo-distributed computing pipelines, especially where mission-critical field operations demand robust edge capabilities—and identify how system architectures must adapt to thrive across this continuum.

System design for modern workloads is end-to-end, e.g., driven by the sheer volume and rate of distributed data that can exceed the capability of on-scene devices. At the same time, high-fidelity data is continuously generated and acquired at the edge by bandwidth-hungry sources, such as scientific instruments, smart manufacturing arrays, and autonomous vehicle fleets. To satisfy strict real-time latency budgets and overcome intermittent network constraints, this raw data must be filtered, compressed, and partially processed near the sensors—often through in-situ feature extraction or low-latency AI inference. Only the highest-value signals and distilled datasets are then transmitted to the cloud, where they are aggregated to fuel resource-intensive tasks like massive-scale deep learning, digital twin simulations, and continuous model refinement.

Stewardship throughout the entire data lifecycle—encompassing data governance, data quality assurance, and policy-driven oversight—has shifted beyond a narrow focus on centralized task processing. Today, it demands sophisticated workflow orchestration and data management across a computing continuum that spans sensors, edge devices, and cloud platforms. At the same time, these use cases are highly performance-critical. They increasingly require capabilities long associated with HPC: efficient parallel execution, high-throughput data movement, high-performance I/O, accelerator utilization, topology-aware communication, and rigorous performance engineering.

# Workshop Format

ECHO will be held as a half-day workshop on the morning of Nov. 20, 2026 (Friday), in conjunction with SC '26 in Chicago, IL. The program combines featured talks that set context, peer-reviewed paper and lightning sessions, a multi-stakeholder panel spanning academia, industry, and national labs, and interactive breakouts for structured ideation and report-backs.

The mix is intentional: guided breakouts drive problem formulation, lightning talks and feedback-oriented Q&A socialize early-stage and work-in-progress results, and mixed panels and breakout groups build bridges across the HPC, edge, distributed-systems, networking, and AI-infrastructure communities.