Close

Presenter

Ru Huang
Peking University
Presentations
Research Manuscript
SpecASR: Accelerating LLM-based Automatic Speech Recognition via Speculative Decoding
2:30pm - 2:45pm PDT Tuesday, June 24 3002, Level 3
AI
AI4: AI/ML System and Platform Design
Research Manuscript
HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference
11:00am - 11:15am PDT Monday, June 23 3001, Level 3
AI
AI4: AI/ML System and Platform Design
Research Manuscript
DuQTTA: Dual Quantized Tensor-Train Adaptation with Decoupling Magnitude-Direction for Efficient Fine-Tuning of LLMs
1:45pm - 2:00pm PDT Monday, June 23 3000, Level 3
AI
AI1: AI/ML Algorithms
Research Manuscript
EdgeMM: Multi-Core CPU with Heterogeneous AI-Extension and Activation-aware Weight Pruning for Multimodal LLMs at Edge
4:00pm - 4:15pm PDT Wednesday, June 25 3002, Level 3
Design
DES1: SoC, Heterogeneous, and Reconfigurable Architectures
Research Manuscript
A Systematic Approach for Multi-Objective Double-Side Clock Tree Synthesis
2:15pm - 2:30pm PDT Wednesday, June 25 3006, Level 3
EDA
EDA7: Physical Design and Verification
Research Manuscript
Local-GS: An Order-Independent Gaussian Splatting Training Accelerator Exploiting Splat Locality
2:30pm - 2:45pm PDT Wednesday, June 25 3000, Level 3
AI
AI3: AI/ML Architecture Design
Research Manuscript
3D-TokSIM: Stacking 3D Memory with Token-Stationary Compute-in-Memory for Speculative LLM Inference
3:30pm - 3:45pm PDT Tuesday, June 24 3000, Level 3
AI
AI3: AI/ML Architecture Design
Research Manuscript
UniCAIM: A Unified CAM/CIM Architecture with Static-Dynamic KV Cache Pruning for Efficient Long-Context LLM Inference
10:30am - 10:45am PDT Monday, June 23 3002, Level 3
Design
DES2A: In-memory and Near-memory Computing Circuits
Research Manuscript
ReaLM: Reliable and Efficient Large Language Model Inference with Statistical Algorithm-Based Fault Tolerance
5:15pm - 5:30pm PDT Wednesday, June 25 3008, Level 3
Systems
SYS6: Time-Critical and Fault-Tolerant System Design