Blog Review: Jun 9

Probabilistic ML; Learning from the past for AVs; AI training; AMBA CHI updates.


Arm’s Partha Maji introduces a collaboration with the University of Cambridge to advance Bayesian statistics and probabilistic machine learning, which could play a vital role in safety-critical AI applications.

Siemens’ Thomas Dewey looks at a way to improve autonomous driving capabilities by enabling vehicles to train on past hazardous situations to provide and early warning for when they might occur in the future.

Cadence’s Paul McLellan finds out what’s new with Habana’s AI training chips for the data center and deployment of Cerebras’ wafer-scale chip.

A Synopsys writer checks out some of the new features in the AMBA CHI protocol, including new request types, transaction flow optimizations, and throughput improvements.

Ansys’ Aaron Edwards points to modeling an entire signal path on a PCB as a way to reveal not only component-level design flaws, but any system-level integration or coupling issues such as signal loss between components.

SEMI’s Sungho Yoon finds that the silicon wafer industry may need to start greenfield projects as soon as this year to boost capacity over the next two years as market demand and average selling prices continue to improve.

Coventor’s Jeonghoon Kim considers how a virtual process library can assist in process module optimization and technology pathfinding.

Nvidia’s Isha Salian checks out a deep learning model that upgrades cosmological simulations from low to high resolution to allow scientists to create a complex simulated universe within a day.

And don’t miss the blogs featured in the latest Automotive, Security & Pervasive Computing and Test, Measurement & Analytics newsletters:

Editor in Chief Ed Sperling examines the impact of new technologies.

Harvard University’s Evelyn Hu reveals the promise of ‘defects’ as quantum mechanical qubits.

Arteris IP’s Kurt Shuler looks at what sets automotive apart from the conventional wisdom on AI hardware markets.

Siemens’ Jacob Wiltgen finds that selecting the right blend of fault injection techniques is important for maximum efficiency.

Xilinx’s Kartik Srinivasan examines the SmartNIC, which combines wired networking and computational resources on the same card to offload tasks from server CPUs.

Synopsys’ Stewart Williams warns that while increasing automotive connectivity brings new opportunities, such as OTA updates, it also raises new risks.

Rambus’ Thierry Kouthon explains why modern vehicle architectures are opening new avenues for attackers and what can be done to keep attackers out.

Flex Logix’s Andy Jaros presents some product architecture choices that make IP integration easier.

Cadence’s Anne-Marie Schelkens demonstrates how CFD is helping to find ways to reduce the noise impact as companies push a return to supersonic flight.

Onto Innovation’s Anoop Somanchi illustrates the role of Fourier transform infrared spectroscopy (FTIR) in monitoring BPSG films in mold layers.

Synopsys’ Pawini Mahajan and Raja Koneru explain how to find issues earlier by validating the connection of clocks and resets.

Advantest’s Koji Miyauchi dissects a test tool that combines conventional ATE and system-level test for OTA test of Bluetooth Low Energy devices.

Leave a Reply

(Note: This name will be displayed publicly)