Chip Industry’s Technical Paper Roundup: October 24

SLAC energy estimates across computing; LLM performance on AI accelerators; LLMs for HW test; power side channel analysis; vertical power delivery; CNNs on embedded automotive systems; in memory computing; multi-bit CAM designs using FeFETs.


New technical papers added to Semiconductor Engineering’s library this week.

Technical Paper Research Organizations
Energy Estimates Across Layers of Computing: From Devices to Large-Scale Applications in Machine Learning for Natural Language Processing, Scientific Computing, and Cryptocurrency Mining SLAC National Laboratory and Stanford University
A Comprehensive Performance Study of Large Language Models on Novel AI Accelerators Argonne National Laboratory, State University of New York, and University of Illinois
LLM4DV: Using Large Language Models for Hardware Test Stimuli Generation University of Cambridge, lowRISC, and Imperial College London
SCAR: Power Side-Channel Analysis at RTL-Level University of Texas at Dallas, Technology Innovation Institute and University of Illinois Chicago
Vertical Power Delivery for Emerging Packaging and Integration Platforms – Power Conversion and Distribution University of Illinois Chicago
Performance/power assessment of CNN packages on embedded automotive platforms University of Modena and Reggio Emilia
First demonstration of in-memory computing crossbar using multi-level Cell FeFET Robert Bosch, University of Stuttgart, Indian Institute of Technology Kanpur, Fraunhofer IPMS, RPTU Kaiserslautern-Landau, and Technical University of Munich
SEE-MCAM: Scalable Multi-bit FeFET Content Addressable Memories for Energy Efficient Associative Search Zhejiang University, China, Georgia Institute of Technology, University of California Irvine, Rochester Institute of Technology, University of Notre Dame, and Laboratory of Collaborative Sensing and Autonomous Unmanned Systems of Zhejiang Province

More Reading
Technical Paper Library home

Leave a Reply

(Note: This name will be displayed publicly)