RESEARCH AREAS

Novel energy-efficient technology-driven computing architectures and platforms, including exploration on Computational Reconfigurable Logic Integrated in Processors, exploration of novel architectures enabled by 3D-IC technologies and non-volatile memory technologies.

Exploration of new energy-efficient applications, such as customized computing for deep learning, image/video processing, mobile healthcare, mobile environment sensing, non-volatile memory in data center, smart sensor networks and Internet of Things.

RESEARCH ACHIEVEMENT

Compilation and Synthesis of Reconfigurable Accelerator-Rich Architecture

Many-Core (GPU/MIC) Architecture and Compiler Optimization

Compilation Techniques for Embedded Systems

Architecture Exploration for Energy-efficient and Reliable Memory Hierarchy

High Performance Storage Systems for Big Data Applications

High-Performance Reconfigurable Architecture for Wireless LAN Physical & Data-link Layers

Design Automation for 3-D Integration and Low-Power Technologies

Energy-efficient Computing Systems for Deep Learning Applications

Acceleration for Medical Imaging

Localization based on Environmental Physical Features and Indoor Floor Plan Reconstruction Techniques

RESEARCH FUNDING

From established to December 2014, CECA has 18 projects supported by national government, and 19 projects supported by industry.

PUBLICATIONS

From established to July 2016, CECA has produced 207 publications at top international conferences and journals in the related research fields, which includes 98 top conference and journal papers.

First time PKU 1st author papers: MobiCom'14, ISCA'15 and MICRO'15 (X2)

First time PKU co-author papers: ASPLOS'14 and ISCA'14