Ashwin Chaugule

Systems Engineer

I work on AI inference infrastructure at Google. Most recently, I led efficiency optimization for Gemini LLM inference, reclaiming 35+% of TPU fleet capacity, and shipped Context Caching on Vertex AI and Google AI Studio from preview to GA.

Previously co-founded ghOSt, a userspace scheduling framework deployed on Google Search (SOSP 2021). Earlier work on Linux memory management solved decade-old kernel problems—including fragmentation-resistant pinning and cache reaper latency (10ms → 50μs).

Before Google: Qualcomm Innovation Center (ARM64 server enablement, upstream Linux kernel). MS Computer Science, Penn State.

Selected Publications

ghOSt: Fast & Flexible User-Space Delegation of Linux Scheduling

J Humphries, N Natu, A Chaugule, O Weisse, B Rhoden, J Don, L Rizzo, O Rombakh, P Turner, C Kozyrakis

SOSP 2021 · ACM SIGOPS Symposium on Operating Systems Principles

Software-Defined Far Memory in Warehouse-Scale Computers

A Lagar-Cavilla, J Ahn, S Souhlal, N Agarwal, R Burny, S Butt, J Chang, A Chaugule, N Deng, J Shahid, G Thelen, K Yurtsever, Y Zhao, P Ranganathan

ASPLOS 2019 · International Conference on Architectural Support for Programming Languages and Operating Systems

Featured by The Morning Paper

A Specification Based Intrusion Detection Framework for Mobile Phones

A Chaugule, Z Xu, S Zhu

ACNS 2011 · International Conference on Applied Cryptography and Network Security

Designing System-Level Defenses against Cellphone Malware

L Xie, X Zhang, A Chaugule, T Jaeger, S Zhu

SRDS 2009 · IEEE International Symposium on Reliable Distributed Systems

ghOSt Scheduler

Kernel and userspace components for delegating Linux scheduling to userspace, enabling rapid iteration on scheduling policies without kernel recompilation.

Kernel Code Userspace Code