Blog Review: May 22

Critical infrastructure vulnerable; speculative execution; diminishing returns for accelerators.


Synopsys’ Taylor Armerding warns that critical infrastructure is still vulnerable to cyber threats, with Kaspersky finding that 42.7% of the industrial control system computers it protected last year were attacked by malware, email phishing, or other threats.

Cadence’s Paul McLellan listens in as Jon Masters of Red Hat considers how to tackle speculative execution and branch prediction vulnerabilities such as Meltdown and Spectre.

Mentor’s Colin Walls shares some embedded software development tips with notes on commenting out code, free software considerations, and deploying a new RTOS.

Arm’s Ilias Vougioukas checks out some current topics in high-performance computing, such as whether accelerators will see diminishing returns in the near future, plus research that crosses computer architecture and biology.

UltraSoC’s Aileen Ryan warns that the market is reacting too slowly to automotive security threats, with cyber-incidents in vehicles in Q1 2019 up more than 300% against the same period last year.

Rambus’ Ben Levine discusses the importance of adequately securing various types of specialized silicon, such as AI accelerators and edge devices.

Walt Custer of Custer Consulting Group looks at trends and history to estimate when the next upturn in the market might happen.

NXP’s Timo van Roermund checks out the upcoming ISO/SAE 21434 standard for cybersecurity in vehicles and how it differs from the ISO 26262 functional safety standard.

Nvidia’s Isha Salian checks out an effort to make Japanese texts written in an obsolete script more accessible and searchable using machine learning.

To switch gears from reading, check out these recently featured videos:

Steven Woo, Rambus fellow and distinguished inventor, explains why choosing memory depends upon data traffic.

Rita Horner, senior technical marketing manager in Synopsys’ Solutions Group, looks at why scalability and cost emerge as top issues with increased data volume.

Ira Leventhal, vice president of Advantest’s new concept product initiative, addresses considerations in using AI to improve reliability in all chips.

Burkhard Huhnke, vice president of automotive at Synopsys, shares tips on how to build and update chips in cars.

Cheng Wang, senior vice president of engineering at Flex Logix, points out where the bottlenecks are in AI chips and how to boost the efficiency.

Ben Levine, senior director of product management for Rambus’ Security Division, explores how interactions between components can compromise AI inferencing models.

Leave a Reply

(Note: This name will be displayed publicly)