Chip Industry Technical Paper Roundup: July 21


New technical papers recently added to Semiconductor Engineering’s library: Technical Paper Research Organizations StreamDQ: Near-Memory Weight DeQuantization in Custom HBM for Scalable AI Inference Acceleration SK hynix Open DRAM Model—Part II: Enabling Processing-in-Memory in 3-D DRAM Georgia Institute of Technology Evaluating Hardware Abstraction Layer Concep... » read more

Monolithic CMOS Platform Integrates Piezo-Optomechanical Photonics (Mitre et al.)


Researchers from MITRE, University of Colorado Boulder, Sandia National Laboratories, University of Arizona, and MIT published a technical paper titled “Monolithic Integration of Piezo-Optomechanical Photonics and CMOS Electronics.” Abstract Excerpt: The paper demonstrates a “fully monolithic, all-CMOS fabricated platform” for piezo-optomechanical photonic integrated circuits and sho... » read more

Research Bits: May 19


Programmable PIC Researchers from the University of Washington designed a low-power programmable photonic integrated circuit that is electrically reconfigurable and can be mass-produced. “This optical chip could help to accelerate the prototyping cycle while reducing power consumption for applications like AI computing. Our study is also the first time someone has shown that these kinds o... » read more

Silicon Photonics: Integrated Grating-Based Antennas for Optical Phased Arrays (MIT)


A new technical paper, "Reduced-crosstalk antennas for grating-lobe-free and wide-field-of-view integrated optical phased arrays," was published by researchers at MIT. Abstract "Integrated optical phased arrays (OPAs) have emerged as a promising technology for many applications due to their ability to dynamically control free-space optical beams in a compact and non-mechanical manner. How... » read more

GPU Power Prediction Tool for AI Workloads (MIT, IBM)


A new technical paper, "EnergAIzer: Fast and Accurate GPU Power Estimation Framework for AI Workloads," was published by researchers at MIT and IBM Research. Abstract "As AI workloads drive increases in datacenter power consumption, accurate GPU power estimation is critical for proactive power management. However, existing power models face a scalability bottleneck not in the modeling tec... » read more

Research Bits: May 5


AI power prediction Researchers from MIT and the MIT-IBM Watson AI Lab developed a prediction tool that can quickly tell data center operators how much power will be consumed by running a particular AI workload on a certain processor or AI accelerator chip. It can be applied to a wide range of hardware configurations. The lightweight estimation model captures the power usage pattern of a GP... » read more

Research Bits: Apr. 28


Parchment papertronics Researchers from Binghamton University used commercial parchment paper, commonly used in baking, along with a standard carbon dioxide laser and water-based conductive ink to create disposable, single-use electronic circuits. The laser selectively removes the paper's thin silicone coating in specific patterns, exposing the water-absorbing cellulose fibers underneath. T... » read more

Why Co-Packaged Optics Should be Viewed as an Architectural Commitment (UW-Madison, MIT et al.)


A new technical paper, "3D optoelectronics and co-packaged optics: when solving the wrong problems stalls deployment," by the University of Wisconsin, MIT, and Invictus Innovation EV Technology. Abstract "The rapid growth of AI and accelerator-driven workloads is forcing a fundamental rethinking of optical interconnect architectures in datacenters. Co-packaged optics and three-dimensional... » read more

Research Bits: Mar. 31


2D hard mask material Researchers from Penn State University and University of Chemistry and Technology Prague propose using the 2D material chromium oxychloride (CrOCl) as a hard mask, because its layered structure is resistant to plasma etching and enables it to be an effective mask at smaller thicknesses. “This 2D material is like lasagna. It’s a layer-by-layer structure,” said Zih... » read more

Research Bits: Mar. 17


Photonic ski jumps Researchers from Massachusetts Institute of Technology (MIT), MITRE, University of Arizona, and Sandia National Laboratories developed a new class of photonic devices that enable the precise broadcasting of light from a chip into free space. The chip uses an array of microscopic structures that curl upward, resembling tiny ski jumps, and allows control over how light is e... » read more

← Older posts