Jie Ye is a PhD candidate in the Department of Computer Science at the Illinois Institute of Technology and a member of the Gnosis Research Center. She is advised by Dr. Xian-He Sun and Dr. Anthony Kougkas. Her research focuses on accelerating LLM/DNN inference; KV cache management and offloading of models and KV caches; advanced data-transfer strategies between training applications and inference systems; optimizing LLM/DNN training; and GPUDirect data transfer.
Joined: Aug 2020Expected: May 2027
16Publications
4Projects
Research Interests
- Distributed Storage
- Parallel and Distributed Systems
- LLM/DNN Inference Optimization
- LLM KV Cache management
- LLM/DNN Training Optimization
Projects
Publications
2026
PKAS: Predictive KVCache-Aware Scheduling for Faster LLM and Transformer Inferences
2026
Pre-RoPE versus Post-RoPE: Trading Per-Step Compute for KVCache Survivability
2026
Accelerating Multi-Agent Orchestration with Speculative Dispatching
2026
Correct Is Not I/O Efficient: Characterizing I/O Behavior of LLM Agents on Scientific Data
2026
PagedEviction: Structured Block-wise KV Cache Pruning for Efficient Large Language Model Inference
2025
Insights into GPUDirect Data Transfer through NIXL Benchmarking
2025
Characterizing the Behavior and Impact of KV Caching on Transformer Inferences under Concurrency
2025
Accelerate LLM inference with Asynchronous model offload
2024
Uncover the Overhead and Resource Usage for Handling KV Cache Overflow in LLM Inference
2024
DaYu: Optimizing Distributed Scientific Workflows by Decoding Dataflow Semantics and Dynamics
2024
HStream: A hierarchical data streaming engine for high-throughput scientific applications
2024
Viper: A High-Performance I/O Framework for Transparently Updating, Storing, and Transferring Deep Neural Network Models
2022
LuxIO: Intelligent Resource Provisioning and Auto-Configuration for Storage Services
2021
HDF5 VOL Connector to Apache Arrow
2021
HFlow: A Dynamic and Elastic Multi-Layered Data Forwarder
2021
