Week In Review: Auto, Security, Pervasive Computing

Apple’s spatial computer; clean hydrogen strategy; V2X collaboration; power electronics for EVs; solid-state battery improvements; power and performance cost of security; CISA’s new vulnerabilities; AI hardware funding.

popularity

Apple uncorked its spatial computer, the Vision Pro, and a new operating system, the visionOS. The “infinite screen real estate” basically untethers the screen from the box, allowing users to work in multiple windows with no space limits. While the device garnered mixed reviews, largely based upon its $3,500 price tag, the implications of mixed-reality computing are potentially significant for the tech industry.

Fig 1: Apple Vision Pro. Source: Apple

Apple also noted that it completed the Mac transition to Apple silicon from Intel-based chips.

The Biden administration uncorked a National Clean Hydrogen Strategy, outlining ways to accelerate the production, processing, delivery, storage, and use of clean hydrogen. Commercial-scale hydrogen could help achieve long-term decarbonization objectives as it produces low or zero carbon emissions.

Meanwhile, Toyota, Hino, Daimler, and Mitsubishi signed a Memorandum of Understanding (MoU) aimed at achieving carbon neutrality, sharing resources to develop hydrogen cars and other technologies, including CASE (Connected/Autonomous & Automated/Shared/Electric). The agreement also formalizes the merger of Hino and Mitsubishi Fuso Truck and Bus Corporation (MFTBC).

Automotive Ecosystem

Infineon Technologies and Autotalks are collaborating on next-generation V2X (Vehicle-to-Everything) applications. V2X lets vehicles communicate with each other and their surroundings to improve road safety. Infineon’s automotive-grade HYPERRAM 3.0 memory supports Autotalks’ SECTON3 V2X and TEKTON3 reference designs for a fully-integrated V2X SoC.

Fig. 2: Automotive-grade HYPERRAM 3.0 memory. Source: Infineon

Electric vehicles require a diverse set of semiconductor technologies, including automotive-grade ICs for battery management and drive train controls, and the ICs and SoCs to manage charging stations along roads and highways. Many of those are based on SiC and GaN, which pose  inspection and test challenges.

Improving battery performance is a key step toward encouraging more people to buy an EV, and scientists at Oak Ridge National Laboratory (ORNL) have found a simple way to achieve this goal. In standard practice, tiny air pockets often block the flow of ions between electrodes when the batteries charge or operate. The new approach creates a material that is almost 1,000 times more conductive. “It’s the same material,” said ORNL lead researcher Marm Dixit. “You’re just changing how you make it, while improving the battery performance on a number of fronts.”

In China, prices of battery-grade lithium carbonate and lithium hydroxide soared, but the price of li-ion batteries didn’t increase, owing to falling prices for contributing materials. The price of EV square ternary cells dropped 9%, energy storage cell prices dropped 12.6%, and consumer battery cell prices fell by 11.5%.

Direct lithium extraction (DLE) is a new streamlined process that increases efficiency and decreases the negative externalities of brine mining. DLE could produce battery-grade lithium carbonate or hydroxide in just a few hours and save the trip to a separate processing facility, making it cheaper and easier to produce batteries for EVs.

Samsung announced that its latest automotive processor, the Exynos Auto V920, has been selected to power Hyundai Motor Company’s next-generation in-vehicle infotainment (IVI) systems, which are expected to roll out by 2025.

Security

For much of the chip industry, concerns about security are relatively new, but the requirement for protecting semiconductor devices is becoming pervasive. However, adding too much protection has a cost. It’s likely to be slower and may consume more power than a device that isn’t as well protected.

Practically unbreakable cybersecurity systems require balancing cost and risk. So even if you could create a system that is “literally” unbreakable, the cost would likely be so high that no one could afford it.

CISA made several announcements this week:

Pervasive Computing

Renesas announced the successful completion of its acquisition of Panthronics, a fabless semiconductor company specializing in high-performance wireless products.

Companies developing AI hardware attracted funding from a range of sources in May.

  • Axelera AI added $23.0M to its Series A round to help design a platform that can provide the AI computational power of a server in a single chip with lower power consumption and price.
  • Zhonghao Xinying Technology raised hundreds of millions of yuan (CNY 100.0M is ~$14.3M) in pre-Series B financing to help design large model AI training chips.
  • Etched.ai raised $5.4M in seed funding to help develop chips specialized for transformer and large language model (LLM) inference.
  • ChipIntelli drew tens of millions of yuan to help develop intelligent voice recognition chips and algorithms for artificial intelligence.
  • NeuronBasic raised tens of millions of yuan to help develop CMOS image sensors (CIS) with integrated image recognition algorithms for low-power edge AI.

Researchers at the University of Texas at Austin have developed a diagnostic device that can tell the difference between influenza and COVID-19. With his team, Ray Chen, professor of computing and electrical energy, built the “lab-on-a-chip” testing technology on a CMOS-compatible photonic chip-based platform. It features a “y”-shaped, double-layered microfluidic chip. The platform can test several samples at once and the aim is to provide more accurate testing solutions and make it easier for people who can’t get to a medical center. In the future, the device may be able to test dozens of illnesses, including other coronaviruses and even some types of cancer.

DARPA chose teams to support its In the Moment program, which aims to support military personnel in challenging situations, such as medical triage, through the use of trusted algorithmic decision-making. Four teams will focus on different technical areas. The teams are Raytheon BBN Technologies and Soar Technology; Kitware and Parallax; CACI International; and the University of Maryland Applied Research Laboratory for Intelligence and Security and the Institute for Defense Analyses.

Research

Recent automotive technical papers:

Click here for more automotive research.

Recent AI/ML/DL technical papers:

  • A PIM Architecture That Supports Floating Point-Precision Computations Within The Memory Chip
  • Comparing Analog And Digital SRAM In-Memory Computing Architectures (KU Leuven)
  • Issues And Opportunities In Using LLMs For Hardware Design

Click here for more AI/ML/DL research.

Recent security technical papers:

  • An Evaluation Of Quantum Algorithms On Classical Hardware Using The CuQuantum Framework
  • Hardware Security: Eliminating/Reducing A Blind Spot Of Side Channels (CISPA Helmholtz Center For Information Security)
  • Rowhammer Vulnerability Of A HBM2 DRAM Chip

Click here for more security papers.

 Events

Find upcoming chip industry events here, including:

  • Radio Frequency Integrated Circuits Symposium-RFIC 2023, June 11 – 13
  • MIPI DevCon 2023: Mobile and Beyond, June 30
  • DAC 2023- Design Automation Conference, July 9 – 13
  • 2023 Flash Memory Conference & Expo, August 8 – 10
  • 32nd USENIX Security Symposium, August 9 – 11
  • SPIE Optics + Photonics 2023, August 20 – 24
  • DARPA: Electronics Resurgence Initiative (ERI), August 22 – 24
  • Hot Chips 2023, August 27 – 29

Upcoming webinars are here.

Further Reading

Read the latest automotive, security, and pervasive computing articles, or check out the latest newsletter.

 



Leave a Reply


(Note: This name will be displayed publicly)