Spring 2026

Elena Petrov, ETH Zurich

Compiler-Guided Locality Management for Disaggregated Memory

1:00 PM · 34-401 · Hybrid

New approaches for orchestrating data placement and movement when memory is no longer tightly coupled to a single compute node.

Marcus Bell, NVIDIA Research

Simulation Infrastructure for Fast Accelerator Design Iteration

1:00 PM · 34-401 · In person

Building simulation stacks that preserve enough fidelity for architecture research without slowing down iteration to a crawl.

Farah Nasser, University of Toronto

Rethinking On-Chip Networks for Multi-Tenant AI Systems

1:00 PM · 34-401 · Zoom

How contention isolation and topology choices interact when accelerators are shared across jobs with different performance goals.

Julian Park, Princeton University

Cross-Layer Power Management for Latency-Critical Services

1:00 PM · 34-401 · In person

Coordinating software control loops and hardware power states to reduce energy use without violating tail-latency targets.