FLex Logix shows off inference chip; OneSpin chips in; Synopsys, IBM collaborate.
Automotive, automation
General Motors is planning a third electric-vehicle plant. The former Saturn factory will make first fully electric Cadillac, in the former Saturn assembly line.
Tesla is allowing some customers to beta test its Full Self-Driving (FSD), according to The Verge. The company pushed the software update to some early access customers to do some real world beta test. Some of the reaction to this push: this is scary, writes The Verge.
SoCs that end up in cars, factories, and medical settings need to be designed to function reliably within a safety, power, performance, and cost. Arm is working with Mentor, a Siemens Business, to offer a new RTL verification design review service for SoC designers to verify the strategy, plan, and closure are correct for the application.
Medical IoT
M3Infekt, a Fraunhofer cluster project — under the lead of Fraunhofer ISS — aims to monitor COVID-19 patients outside of the hospital to flag distressing symptoms earlier when a patient takes a turn for the worst. The 3 M’s of the project’s name are modular, multimodal and mobile system.
Medical imaging company Butterfly Network won the Ultrasonic Transducers Product Showdown at SEMI’s MEMS and Sensors Executive Congress for its Butterfly iQ product. The iQ is a handheld ultrasound systems connected via IoT that lowers the cost of medical imaging by 10 times. TDK-Chirp Microsystems and Ultrasense Systems were finalists.
Edge security
OneSpin is contributing security to German government-funded project to build a secure and scalable ecosystem for a RISC-V-based systems for AI edge systems more straightforward and secure. As a contributing partner in the Scalable Infrastructure for Edge Computing (Scale4Edge) project, OneSpin is contributing verification tools.
Edge AI
Flex Logix announced its edge AI inference accelerator, the InferX X1 chip, at the Linley Fall Processor Conference this week. The chip is designed to be small for the edge uses — silicon area is 54mm2 and is much smaller than NVIDIA’s Jetson Xavier at 350mm2 — and still maintains high-quality, high performance, says a press release from Flex Logix. InferX X1 is not made to prune the data because customers do not want that, but customers have the flexibility to build in some pruning. InferX X1 mass production chips and software will be available Q2 2021.
Arm added the Ethos-U65 microNPU (Neural Processing Unit) to its line of machine learning/ AI processers. The Ethos-U65 is power efficient and now supports Arm Cortex-A and Arm Neoverse-based systems.
Data centers, cloud
Intel has agreed to sell storage unit to SK Hynix for $9 billion, to free up Intel to concentrate on its core business. The deal does not include the Optane division.
IBM used Synopsys tools in IBM Research’s AI Hardware Center to advance the development of chip architectures and design methodologies critical to the next generation of AI chips, said Synopsys in a press release.
Company milestones and wins
Aldec added support for Xilinx PYNQ (Python Productivity for Zynq) into the TySOM embedded system development tools. The high-level programing language Python is open source, which should help put design of the Xilinx SoC and MPSoC devices into a software programmers skill set.
Synopsys’s 3DIC Compiler recently made it possible for Samsung Foundry to design, implement and tape out a complex 5-nanometer SoC with eight high-bandwidth memories (HBMs) in a single package.
Job, Event and Webinar Boards: Find industry jobs and upcoming conferences and webinars all in one place on Semiconductor Engineering. Knowledge Center: Boost your semiconductor industry knowledge. Videos: See the latest Semiconductor Engineering videos.
Leave a Reply