Arm Total Compute 2023; Tenstorrent, LG AI chips; DOT $52M on smart roads.
AI predictions and announcements filled the news this week, including a statement from the Center for AI Safety that was signed by some top AI execs — including Sam Altman, CEO of OpenAI — warning that uncontrolled AI could end up smarter than us and lead to our extinction. Foxconn estimates its artificial intelligence server revenue will double this year with the popularity of generative AI.
Tenstorrent, an AI chip startup headquartered in Canada, is working with LG Electronics to design and produce new RISC-V, AI, and video codec chiplets for LG’s TV and automotive products and Tenstorrent’s data center products. Tenstorrent’s AI and RISC-V CPU technology with LG’s video codec technology is planned to yield chiplet and IP for improving video processing in Tenstorrent’s future data center products.
The U.S. Department of Transportation’s (DOT) National Highway Traffic Safety Administration (NHTSA) is proposing a rule to require automatic emergency braking (AEB) and pedestrian AEB (PAEB) systems on passenger cars and light trucks. NHTSA issued its notice of proposed rulemaking (NPRM) and is soliciting comments. When finalized, the rule will become part of the Federal Motor Vehicle Safety Standard (FMVSS).
The U.S. DOT also announced it approved $52 million in grants to technology-first programs to improve road safety in states, including $1 million for Arizona to install 50 vehicle-to-infrastructure (V2I) traffic control sign systems in rural Mohave County and $5 million to Utah for a statewide expansion of its Connected Communities V2X program. Almost $12 million goes to Maryland to put sensors and other new technologies along a 113-mile-stretch of Route 50 for traffic warnings, monitoring, and predictions. The University of Michigan Transportation Research Institute (UMTRI) will receive $9.8 million for cellular vehicle-to-everything (C-V2X) research and deployments, and Minnesota will receive $9.3 million to expand its Autonomous Rural Transit Initiative (MARTI) — an on-demand automated micro-transit service. The projects are part of the Advanced Transportation Technology and Innovation (ATTAIN) program.
Tesla’s Elon Musk and Nvidia’s Jensen Huang made headlines with their respective visits to China and Taiwan. Musk met with China’s ruling party officials and was reportedly feted for Tesla’s EV presence in China. Huang was treated like a rock star, according to Bloomberg, and followed around by fans. He made statements warning about underestimating China’s ability to catch up in chip manufacturing, and how that will be bad for established chip companies, and said he felt safe relying on Taiwan for chip manufacturing. Nvidia unveiled generative AI platforms powered by its Grace Hopper superchips at Computex in Taipei, and Huang will visit China to meet tech executives.
SoftBank Corp. will be using Nvidia’s Arm-based Grace Hopper superchip for new, distributed AI data centers across Japan.
Arm announced its mobile computing platform Arm Total Compute Solutions 2023 (TCS23), for premium solution for smartphones. The platform has IP for specific workloads, the Arm Immortalis GPU on Arm’s 5th Generation GPU architecture for visual display, the new cluster of Armv9 CPUs for AI, and more accessible software. Cadence Design Systems recently developed RTL-to-GDS digital flow Rapid Adoption Kits for the new Arm TCS23, using its AI-driven Cadence Cerebrus to automate and scale digital chip design. Synopsys also announced its AI-design capabilities for Arm.
Northern Rangelands Trust (NRT) and Connected Conservation Foundation (CCF) set up an IoT (Internet of Things) network in Kenya preserves to watch over range land and catch poachers. The LoRaWAN (Long Range Wide Area Network), made up of LoRaWAN IoT sensors and managed with Actility’s ThingPark platform, is according to the organizations, Africa’s largest landscape-wide IoT conservation network.
Renesas uncorked three new MCU groups targeting motor control. There are a total of 35 new devices, some from RA and RX groups, including MCU and MPU families, analog and power solutions, sensors, communications devices, and signal conditioners, among others.
Keysight and Nokia Bell Labs will demo radio frequency integrated circuit (RFIC) characterization for 6G subterahertz (sub-THz) RF devices in the D- and E-bands at the International Microwave Symposium 2023 (IMS 2023) in June.
Keysight added a lidar target simulator to its test equipment portfolio for autonomous driving. The E8717A Lidar Target Simulator (LTS) simulates test targets in a compact, standardized bench setup. The targets have simulated target distances from 3 meters to 300 meters and surface reflectivity from 10% to 94%. A cobot helps with test set up.
MediaTek and Nvidia are working together on AI-based automotive intelligent cabins for software-defined, connected cars. MediaTek will integrate Nvidia’s Arm-based GPU chiplet and Nvidia’s Drive into its automotive SoC, running the NVIDIA DRIVE OS, DRIVE IX, CUDA, and TensorRT software technologies on the new SoC. Nvidia will offer a new GPU chiplet with some MediaTek technology to an automotive market, while MediaTek will be able to add beef up its Dimensity Auto with new cockpit and in-cabin graphics capabilities from Nvidia.
Stellantis says it will have an all-electric SUV in 2025 and is investing 160 million euros ($176.13 million) in their production. To get prepared, Stellantis has been investing in batteries. Owners/investors Stellantis, TotalEnergies, and Mercedes-Benz celebrated the scaling up of high-performance battery production at their first Automotive Cells Company (ACC) gigafactory in Billy-Berclau Douvrin, France, one of three such battery plants planned in Europe. The plant’s initial production line capacity will be 13 gigawatt-hours (GWh), rising to 40GWh by 2030. Stellantis’ goal is to have 400 GWh of capacity by 2030, through five gigafactories in Europe and North America, in addition to supply contracts. Also, Stellantis and Vulcan Energy, an Australian lithium company, signed an agreement to work on a geothermal energy source in France for a Stellantis factory.
Among the challenges facing autonomous driving design today are the limitations of different sensors. To achieve safe autonomous driving, sensor fusion may be needed. The key questions are not only how many sensors, what types, and where to deploy them, but also how AI/ML technology should interact with sensors to analyze data for optimal driving decision making.
The U.S. Federal Trade Commission (FTC) and the Department of Justice (DOJ) say Amazon, the owner of Ring, did not protect its customers from hackers, who gained access to Ring camera data (via ‘credential stuffing’ and ‘brute force’ attacks), and an employee who spied on customers. Ring must delete data, models, and algorithms derived from videos it unlawfully reviewed. “Ring’s disregard for privacy and security exposed consumers to spying and harassment,” said Samuel Levine, director of the FTC’s Bureau of Consumer Protection, in a press release. “The FTC’s order makes clear that putting profit over privacy doesn’t pay.” Also Amazon will pay $25 million to customers whose Alexa data — some of which were sound recordings of children — was not deleted by Amazon when requested. The FTC and the DOJ have charged Amazon with violating children’s privacy law.
SEALSQ, a company that develops certified secure microcontrollers and implementing post-quantum cryptography, PKI, and post-quantum technology hardware and software products, has been listed on the Nasdaq Global Market as LAES. A subsidiary of SEALSQ makes post-quantum microchips and devices that can be used in a variety of applications.
The security of AI/ML systems and chips made with AI/ML is a major concern. AI/ML systems are moving targets. Algorithms are being changed, EDA tools are incorporating, and chips themselves are disaggregating into component parts that may or may not include some form of AI/ML/DL. The result is much greater potential for mischief, for IP theft, and for frontal and side-channel attacks that can corrupt data or lead to ransomware. Many of AI/ML systems are vulnerable, and the companies using this technology have little understanding of how to fix problems when they arise.
Recent automotive technical papers:
Click here for more automotive research.
Recent AI/ML/DL technical paper:
Click here for more AI/ML/DL research.
Recent security technical papers:
Click here for more security papers.
Security
See more events on our events page.
Read the latest automotive, security, and pervasive computing articles, or check out the latest newsletter. We update the article and newsletter pages throughout the month with new content.
 |
 |
 |
 |
 |
 |
Leave a Reply