News

Mark is teaching CSCI 5942: AI Engineering this fall.

Strata: Hierarchical Context Caching for Long Context Language Model Serving appeared at OSDI 2026.

SYMI: Efficient Mixture-of-Experts Training via Model and Optimizer State Decoupling appeared at NSDI 2026.

SHARD: A Compatibility Framework for Deploying Transformer Models on Edge NPUs appeared at EuroMLSys 2026.

vCache: Verified Semantic Prompt Caching appeared at ICLR 2026.

Beyond the Accelerator: A Full-Stack HW/SW Co-Design Analysis for Recommendation System Inference appeared in IEEE Micro.

cedar: Optimized and Unified Machine Learning Input Data Pipelines appeared at VLDB 2025.

BASIL launches at CU Boulder. We’re recruiting students at all levels.