Blog Review: April 7

Semi manufacturing and national security; more Python verification; PCIe 6.0 overcomes bandwidth limitations; edge image classification.


Cadence’s Paul McLellan checks out the US National Security Commission on Artificial Intelligence report and what it recommends for funding the development of AI as well as semiconductor manufacturing and research.

Siemens EDA’s Ray Salemi continues exploring Python for verification and shows how to use cocotb to create a simple bus functional model and connect it to a testbench.

Synopsys’ Priyank Shukla checks out how the upcoming PCIe 6.0 specification can help solve the bandwidth limitations that HPC SoCs are constantly facing through faster data transfer and the move to a PAM-4 architecture.

Arm’s Masoud Koleini demonstrates how to implement an image classification system that runs machine learning inference over a stream of data on a low-power edge NVIDIA Jetson Nano device using RedisAI.

Ansys’ Nikhil Grover offers some simple guidelines for electromagnetic compatibility when designing PCBs and how automatic and customizable signal reference rules checks of PCBs can identify areas of potential issues.

In a blog for SEMI, Tarek Zohdi, Simo A. Mäkiharju, Evan Variano, and Pushkar Apte of UC Berkeley simulate and analyze particle and droplet transport, brought to attention during COVID-19, to enable semiconductor manufacturing cleanrooms to operate more safely.

For more good reading, check out the blogs featured in the latest Automotive, Security & Pervasive Computing and Test, Measurement & Analytics newsletters:

Flex Logix’s Vinay Mehta offers simple steps to make sure you get the fastest inferences.

Synopsys’ Dana Neustadter explains why high-speed interfaces are getting new security requirements to better protect sensitive data and communications.

Arteris IP’s Kurt Shuler looks at how even small IoT designs can have plenty of complexity in architecture and integration.

Siemens’ Ahmed Majeed Khan lays out the security strategies OEMs can use to protect vehicle entry points and in-vehicle networks.

Rambus’ Maxim Demchenko advises using security anchored in hardware at the communication layer to protect data in motion.

Cadence’s Paul McLellan sketches out why understanding the business models of cybercrime points to ways to defend against it.

Onto’s Will Zhou explains how neural networks perform defect classification and many other tasks.

Synopsys’ Taylor Armerding examines the effort to establish automotive cybersecurity standards and best practices.

Advantest’s Matthias Stahl lays out ways of keeping up with the test requirements of a new generation of devices.

Calibra’s Jan Willis summarizes TSMC’s Danping Peng’s take on the historical hurdles for curvilinear technology.

Leave a Reply

(Note: This name will be displayed publicly)