Sotabase

Career

· Research Intern, Meta2025–
· PhD in Computer Science, University of Washington2021–2026

Publications (14)

POD-Attention: Unlocking Full Prefill-Decode Overlap for Faster LLM Inference
International Conference on Architectural Support for Programming Languages and Operating Systems · 2024
47
cited
GPM: leveraging persistent memory from a GPU
International Conference on Architectural Support for Programming Languages and Operating Systems · 2022
19
cited
ScoRD: A Scoped Race Detector for GPUs
International Symposium on Computer Architecture · 2020
19
cited
iGUARD: In-GPU Advanced Race Detection
Symposium on Operating Systems Principles · 2021
17
cited
LLeMpower: Understanding Disparities in the Control and Access of Large Language Models
arXiv.org · 2024
8
cited
Scalable, Programmable and Dense: The HammerBlade Open-Source RISC-V Manycore
International Symposium on Computer Architecture · 2024
6
cited
Scoped Buffered Persistency Model for GPUs
International Conference on Architectural Support for Programming Languages and Operating Systems · 2023
6
cited
Herding LLaMaS: Using LLMs as an OS Module
arXiv.org · 2024
4
cited
MaxMem: Colocation and Performance for Big Data Applications on Tiered Main Memory Servers
arXiv.org · 2023
4
cited
(MC)2: Lazy MemCopy at the Memory Controller
International Symposium on Computer Architecture · 2024
4
cited
Storage Class Memory: Principles, Problems, and Possibilities
arXiv.org · 2019
4
cited
LLM-42: Enabling Determinism in LLM Inference with Verified Speculation
2026
1
cited
Sobriety Testing Based on Thermal Infrared Images Using Convolutional Neural Networks
IEEE Region 10 Conference · 2018
1
cited
Achieving Guaranteed Performance for Protection Traffic in Smart Grid Wide-Area Networks
2019 IEEE 7th International Conference on Smart Energy Grid Engineering (SEGE) · 2019
Sotabase
Aditya K Kamath | Researcher Profile | Sotabase | Sotabase