Blog Review: Feb. 14

Automated reliability verification; tricking neural nets; blockchain boom; simulating satellites.


Mentor’s Matthew Hogan takes a look at why it’s important to establish a baseline reliability verification process and how foundry rule decks fit in.

Synopsys’ Robert Vamosi digs into the issues with fitness tracker Strava’s heatmap, how it could be manipulated, and why the risks of big data analytics go beyond wearables.

Cadence’s Paul McLellan points to research showing how easy it can be to trick neural networks with carefully-added noise, with potentially hazardous results.

Rambus’ Aharon Etengoff notes that spending on blockchain technologies is set to double to $2.1 billion in 2018 as companies experiment with new use cases and the U.S. government investigates applications.

Arm’s Gian Marco Iodice provides a tutorial on implementing the convolutional neural network AlexNet on a Raspberry Pi using the Compute Library.

Ansys’ Paolo Colombo argues that while we’ve become used to massive, easy connectivity, we are just at the beginning of a communications revolution, and space satellites will play a big role.

Cadence’s Meera Collier gives an overview of ISO 26262 and determining how safe is safe enough when it comes to automotive.

In a podcast, Mentor’s John McMillan discusses key areas in PCB design that can help accelerate product design and launch to market.

Synopsys’ Iain Singleton considers what unpopped popcorn and corner cases have in common and how formal can help.

Plus, check out the highlighted blogs from last week’s Low Power-High Performance newsletter:

Editor In Chief Ed Sperling points to why it’s becoming more difficult to achieve the usual balance in designs.

Fraunhofer’s Andy Heinig explains why complex packages make it necessary to handle data in a machine-readable format.

Mentor’s Derong Yan digs into verification methods and how they can help with electrostatic discharge.

Moortec’s Ramsay Allen considers the effects and mechanisms of chip aging, plus how to predict a device’s lifetime.

Rambus’ Frank Ferro argues that the path to GDDR memory isn’t so clear.

Cadence’s Larry Przywara shows how DSPs enable a new ubiquitous interface.

Synopsys’ Licinio Sousa and Arm’s Vassilis Androutsopoulos peel back the curtains on display subsystem architectures for high-end smartphones.

Arm’s John Ronco contends that proven architectures can energize the long tail of the IoT market.

Leave a Reply

(Note: This name will be displayed publicly)