11 papers · 1 filter
Rethinking Unified Memory for NPU-PIM Systems: Dual-View Memory for Dynamic Inference of LLM
Shixin Zhao, Lian Liu, Tianhua Han +3
Heterogeneous architectures that combine neural processing unit (NPU) and processing-in-memory (PIM) are increasingly adopted to accelerate LLM inference. Prior work focuses on bui…
Ouroboros: Wafer-Scale SRAM CIM with Token-Grained Pipelining for Large Language Model Inference
Yiqi Liu, Yudong Pan, Mengdi Wang +5
Conventional LLM inference architectures suffer from high energy and latency due to frequent data movement across memory hierarchies. We propose Ouroboros, a wafer-scale SRAM-based…
TriMoE: Augmenting GPU with AMX-Enabled CPU and DIMM-NDP for High-Throughput MoE Inference via Offloading
Yudong Pan, Yintao He, Tianhua Han +7
To deploy large Mixture-of-Experts (MoE) models cost-effectively, offloading-based single-GPU heterogeneous inference is crucial. While GPU-CPU architectures that offload cold expe…
PAM: Processing Across Memory Hierarchy for Efficient KV-centric LLM Serving System
Lian Liu, Shixin Zhao, Yutian Zhou +4
The widespread adoption of Large Language Models (LLMs) has exponentially increased the demand for efficient serving systems. With growing requests and context lengths, key-value (…
From Buffers to Registers: Unlocking Fine-Grained FlashAttention with Hybrid-Bonded 3D NPU Co-Design
Jinxin Yu, Yudong Pan, Mengdi Wang +4
Transformer-based models dominate modern AI workloads but exacerbate memory bottlenecks due to their quadratic attention complexity and ever-growing model sizes. Existing accelerat…
Lifecycle Cost-Effectiveness Modeling for Redundancy-Enhanced Multi-Chiplet Architectures
Zizhen Liu, Fangzhiyi Wang, Mengdi Wang +4
The growing demand for compute-intensive applications has made multi-chiplet architectures a promising alternative to monolithic designs, offering improved scalability and manufact…