WeightBridge, a weight-transfer library for reinforcement learning systems, reduces average GPU stall time by up to 42× over the state-of-the-art open-source RL framework across configurations spannin…
Researchers Weifan Jiang and colleagues at Harvard University proposed H-Spec, a hybrid Mamba-attention parallel speculative decoding drafter that eliminates the drafter-side KV cache by reusing targe…
Anshvardhan Shetty, a first-year EECS undergraduate at Imperial College London and research intern with Professor Juncheng Yang at Harvard SEAS, presents a study of KV cache admission on high-bandwidt…
RouteWise, a multi-provider LLM routing system presented by Haoran Ni of Nanjing University, reduces cost by 26.6%, mean time to first token by 58.4%, and SLO violations by 98.2% compared with OpenRou…
In their OSDI '26 paper, researchers propose Learning-Augmented Heuristics (LAH), a cache eviction design that keeps a simple heuristic on the request path and uses machine learning to periodically co…
Zishen Wan, an incoming assistant professor at Columbia CS (Fall 2026) and currently a postdoc at Harvard, will present his research on AI-native computing systems, covering workload characterization …
Akira van de Groenendaal, a senior studying Computer Science at Carnegie Mellon University, presented research at the Harvard Systems Group showing that bursty workloads can improve the time-per-outpu…
A benchmark study by an independent researcher found that burstier request arrivals speed up LLM inference, contradicting standard intuition. The analysis of vLLM serving shows that higher burstiness …
MorphServe, a system developed by researchers and detailed in a paper on arXiv, enables elastic model precision for bursty LLM serving by temporarily swapping low-impact FP16 layers to INT4 quantizati…
Saurabh Bagchi, a professor at Purdue University and director of the NSF CISE Center Chorus, will present a talk examining AI's dual impact on computing system reliability and security, highlighting b…
Minghao Li, a fifth-year PhD student in Computer Science at Harvard SEAS, presented a series of works at the intersection of networking and machine learning systems, arguing that distributed ML system…