NXP Semiconductors has introduced a new automotive radio and audio processor that significantly enhances artificial intelligence (AI)-based audio processing, targeting next-generation software-defined vehicles.

The company announced on July 21 that it has launched the SAF9800, an automotive radio and audio processor that integrates analog radio reception and AI/ML-enabled audio processing into a single chip.

The SAF9800 combines AM/FM radio reception with an integrated audio digital signal processor (DSP).

Its DSP supports AI-based sound separation, enabling the system to distinguish voice commands, emergency vehicle sirens and abnormal vehicle noises from surrounding background noise. The processor also provides real-time noise reduction and adaptive audio processing capabilities.

NXP said the chip integrates a neural network-enabled DSP and dedicated hardware accelerators, delivering more than 10 times the audio processing performance of the previous generation.

The processor also incorporates EVAM-lite, NXP's latest AM radio denoising technology. Electric vehicles often experience degraded AM radio reception because electromagnetic interference generated by electric motors and inverters affects AM frequencies. According to NXP, EVAM-lite helps mitigate this interference. The technology comes as US lawmakers continue to consider legislation requiring AM radio in new vehicles because of its role in delivering emergency alerts.

NXP said automakers using the SAF9800 will be able to configure and update radio and audio functions through software, responding to the shift toward software-defined vehicles (SDV).

The SAF9800 began supplying customers starting in May 2026.

"The automotive industry is transitioning from hardware-centric infotainment systems to integrated software-defined architectures," said John van den Braak, senior vice president of Infotainment Audio Systems at NXP. "With a flexible single-chip solution, we aim to simplify vehicle development while enabling automakers to scale their platforms."

Infineon Technologies has introduced the RIC70115, a radiation-hardened gate driver designed to control gallium nitride (GaN) power semiconductors in space applications, the company announced on July 15.

A gate driver is a control IC that switches power semiconductors on and off. The RIC70115 supports both silicon and GaN power devices and is engineered to withstand radiation-induced performance degradation and malfunction, making it suitable for satellites and other space systems, according to the company.

Prolonged exposure to radiation in space can gradually degrade semiconductor performance, while high-energy particles may trigger transient circuit failures.

Infineon said the RIC70115 is qualified to withstand a total ionizing dose (TID) of up to 100 krad(Si) and has been verified to tolerate single-event effects up to a linear energy transfer (LET) level of 81.9 MeV·cm²/mg.

The company said these specifications provide sufficient radiation tolerance for low Earth orbit (LEO) satellites as well as more demanding space environments.

The RIC70115 incorporates a Miller clamp function that prevents unintended turn-on of power semiconductors caused by voltage transients. It also includes integrated noise reduction circuitry to minimize electromagnetic interference.

An integrated low-dropout (LDO) regulator generates a 4.8V gate drive supply from either a 5V or 12V input, reducing the number of required external components.

The device operates across a temperature range of -55°C to 125°C and complies with MIL-PRF-38535, the US military specification for high-reliability microelectronics used in defense and space applications.

"Adoption of GaN power solutions is accelerating rapidly," said Mike Mills, senior vice president of Infineon's High Reliability (HiRel) business. "The RIC70115 is designed to support the transition to GaN-based power systems for space applications."

The RIC70115 and its evaluation board, RIC70115EVAL1, are now commercially available.

Rapidus is bringing agentic artificial intelligence into its foundry platform as it seeks to attract customers for its upcoming 2-nanometer manufacturing process.

The Japanese foundry announced on July 16 that it is partnering with US electronic design automation (EDA) software provider Cadence Design Systems to advance AI-assisted chip design.

Under the collaboration, Cadence's InnoStack AI Super Agent will be integrated into RAADS (Rapidus AI-Agentic Design Solution), the company's chip design platform.

The partnership is believed to mark the first time a foundry has directly integrated an EDA vendor's agentic AI technology into its customer-facing design platform. Industry observers view the move as part of Rapidus' strategy to strengthen its competitiveness and attract customers as a late entrant to the global foundry market.

Unlike conventional AI tools that optimize individual design tasks, agentic AI can independently determine which design tools to use, execute them and decide the next steps without requiring engineers to issue commands for each stage. The technology is designed to orchestrate the entire chip design workflow rather than optimize isolated tasks.

Cadence's InnoStack automates the execution of conventional EDA tools used during the back-end stages of chip design, including placement and routing, timing analysis and power optimization. Instead of engineers repeatedly adjusting design parameters manually, the AI explores design alternatives and identifies optimized solutions.

Rapidus said the collaboration is expected to reduce the time required from initial chip design through final pre-manufacturing verification by as much as 50%.

Rapidus is backed by major Japanese companies including Toyota Motor, Sony Group, SoftBank, Kioxia, and Nippon Telegraph and Telephone (NTT), with substantial financial support from the Japanese government. The company is developing a 2nm process based on technology from IBM and aims to begin mass production in 2027.

"Advanced system-on-chip (SoC) design requires AI agents that coordinate the entire design flow, not AI that simply optimizes individual tasks," said Anirudh Devgan, president and chief executive officer of Cadence. "Our technology will help customers improve productivity and bring advanced chips to market more quickly."

Atsuyoshi Koike, representative director and chief executive officer of Rapidus, said combining RAADS with InnoStack will enable customers to design increasingly complex SoCs more efficiently.

"We expect the integration to significantly reduce design time while maximizing the value of Rapidus' advanced process technologies," Koike said.

Knowles, the US company best known as a supplier of microelectromechanical systems (MEMS) microphones for Apple's iPhone, is expanding its business into aerospace, defense, industrial and medical applications as it completes its transition away from consumer electronics.

On July 17, the company announced that it is expanding its portfolio of high-performance pulse power capacitors. Pulse power technology stores electrical energy and releases it in a highly controlled burst over an extremely short period. Capacitors are key components that store electricity and discharge it precisely when needed.

The company's pulse power capacitors are designed for applications including defense radar and electronic warfare systems, CO2 lasers used in semiconductor manufacturing, drilling equipment for energy exploration and medical defibrillators. These systems require highly reliable capacitors capable of delivering precise amounts of energy repeatedly at exactly the right moment.

"Consistency is everything in pulse power," said Jeff Niew, president and chief executive officer of Knowles. "Once a system leaves the laboratory and enters real-world applications, components must do more than work once—they must deliver stable, repeatable performance under demanding operating conditions."

Founded in 1946, Knowles built its reputation as a supplier of miniature microphones and receivers for hearing aids before becoming a leading manufacturer of MEMS microphones for smartphones during the 2000s. The company was spun off from US industrial manufacturer Dover Corporation in 2014 and subsequently listed on the New York Stock Exchange.

Knowles began reshaping its business in 2023. In September of that year, it announced a strategic transformation into an industrial technology company and acquired US film capacitor manufacturer Cornell Dubilier for $263 million, expanding its portfolio beyond ceramic capacitors into high-power film capacitors for industrial applications. The newly introduced pulse power capacitor lineup is based on technologies acquired through that transaction.

In 2024, Knowles sold its consumer MEMS microphone business, which supplied components for smartphones and wireless earbuds, to US AI semiconductor company Syntiant for $150 million. The company said the divestiture reflected its strategy to reduce exposure to consumer electronics and concentrate on high-reliability components for defense, medical and industrial markets.

The restructuring has begun to improve financial performance. Knowles reported first-quarter revenue of $153.1 million, up 15.8% from a year earlier. Its Precision Devices segment, which includes capacitor and radio-frequency (RF) components, recorded new orders exceeding revenue by 19% during the quarter, marking the sixth consecutive quarter in which orders outpaced sales. The company subsequently raised its full-year financial guidance.

Today, Knowles operates two core businesses: medical acoustic components, including products for hearing aids, and precision capacitor and RF components.

OpenAI is reportedly preparing to enter the consumer hardware market with a portable, screen-free smart speaker designed to serve as an AI companion in the home.

Sources told Bloomberg the device will access ChatGPT’s AI capabilities to control smart home appliances, play media, answer questions and respond to messages. It will utilise OpenAI’s GPT-Live voice technology to listen and adaptively respond at the same time to support more natural conversations and incorporate AI models beyond those used in existing smart speakers.

In addition, the hardware will apparently feature a camera and sensors to enable the AI to read its surroundings. Unlike conventional smart speakers, the rechargeable device will be portable and feature moving mechanical elements.

It will also draw on user data such as emails to become increasingly personalised over time. OpenAI envisions the speaker developing a humanlike personality, proactively surfacing information, anticipating users’ needs and acting as a physical embodiment of ChatGPT.

OpenAI’s hardware unit is reportedly developing around five products in total, with the speaker set to be the first to market.

Last year, the company splashed $6.5 billion to acquire io Products, the device startup co-founded by former Apple design chief Jony Ive. Ive’s creative design studio LoveFrom is also working on the hardware line.

A hardware launch would put the ChatGPT-maker in direct competition with Apple, Amazon and Google in consumer devices.

Trade secrets
Last week, Apple filed a lawsuit against OpenAI and io Products, alleging “theft of trade secrets” obtained through former employees during hiring to accelerate the development of its AI devices. According to its lawsuit, more than 400 former Apple employees have joined OpenAI’s hardware effort.

Drew Pusateri, a spokesman for OpenAI, told the BBC OpenAI has “no interest in other companies’ trade secrets” and is “focused on building innovative technology that empowers people everywhere”.

Apple is seeking an injunction that could delay the AI giant’s device sales. According to Bloomberg, OpenAI is aiming to unveil the speaker later this year, with release targeted for 2027, subject to the lawsuit.

Meanwhile, Apple is reportedly developing its own AI-focused home devices including a smart home hub with a 7-inch display and facial recognition tied to the upcoming AI-powered Siri revamp in iOS 27.

The post OpenAI plots AI speaker launch in hardware push appeared first on Mobile World Live.

The UK’s Competition Appeal Tribunal approved a £3 billion class action lawsuit against Apple over its iCloud service, allowing consumer group Which? to proceed with a claim on behalf of nearly 40 million users across the country.

The lawsuit alleges Apple abused its dominant position and breached UK competition law by “trapping” users into using iCloud. Which? argued Apple steered iOS users towards its own cloud storage service and limited visibility of alternative providers, leading to “rip-off” prices and lower levels of free storage.

The case is expected to go to trial in 2028 after the UK court granted Which? a Collective Proceedings Order (CPO), clearing the path for what could become one of the UK’s largest consumer tech claims.

At the centre of the dispute is Apple’s cloud storage model, where users receive 5GB of free storage before being directed towards paid tiers. Pricing then rises across multiple levels, from 99p per month for 50GB to £54.99 for 12TB in the UK. The consumer group argued this structure reflects a lack of competition, with users “locked-in” once their devices become reliant on iCloud for photos, messages and device data backups.

Strong message
The claim, first filed in November 2024, will cover consumers living in the UK who used iCloud services between 8 November 2018 and 8 June 2026. Potential compensation is estimated at around £77 per user if it succeeds.

Which? also called on the tech giant to resolve the dispute without litigation, urging the company to offer consumers refunds and open up iOS “to allow users a real choice for cloud services”.

However, Reuters reported Apple previously rejected the allegations, arguing users are not required to use iCloud and that third-party alternatives remain available.

Which? CEO Anabel Hoult said the group “wants to make clear that no company, no matter how powerful, can get away with abusing its position”. “The granting of the CPO means we’re one step closer to getting consumers the redress we believe they are owed from Apple,” she noted, adding, “this should send a strong message to any other companies using anti-competitive tactics”.

The post UK court greenlights £3B Apple iCloud lawsuit appeared first on Mobile World Live.

Telia identified emergency services, energy, transport, industry and healthcare as target sectors for an IoT connectivity service using standalone (SA) 5G the company launched in Sweden.

The operator asserted Telia Critical IoT Connectivity was the first commercially available product of its type in the country to use the network technology, highlighting it had already rolled out SA 5G widely in the nation.

Pointing to frequently promoted advantages of the network architecture for slicing and provision of low-latency, secure communications, it noted SA 5G enabled smart traffic management, and is optimised for video and data.

These factors ensure “reliable performance for key applications even at peak times”.

Head of B2B at Telia Sweden Fredrik Stenberg said the product creates the conditions for customers to “use new digital solutions in the most critical parts of their operations, where the demands are greatest, without compromising security or control”.

The service has been developed partly using the experience and insights Telia gained from NorthStar 5G, a programme launched alongside Ericsson carrying European Union funding which is focused on increasing uptake of the technology in industry.

Its commercial offer on SA 5G is Telia’s second major IoT product announcement in the space of a month.

In May it unveiled Telia IoT Connect, a service providing sovereign infrastructure aimed at companies and authorities requiring secure and resilient communications while keeping all data within Sweden.

The post Telia employs SA 5G for critical IoT appeared first on Mobile World Live.

AT&T unveiled a 24-hour unlimited wireless data service for iPad users in the US, available to any customer regardless of their existing operator for a flat daily rate of $3.

The company claimed to be the first major US wireless provider to offer on-demand connectivity for iPad users on a pay-as-you-go basis. The service is for models with mobile and Wi-Fi connectivity capabilities, with the 24-hour data window beginning shortly after purchase.

AT&T explained it is targeting consumers who do not activate mobile connectivity on compatible iPads, pitching the benefit of short-term network access for periods where users are not in Wi-Fi range.

It highlighted ease of use, stating the service is activated in device settings without the need for a dedicated app.

The first day pass is complimentary, with subsequent purchases charged to a credit or debit card.

Josh Goodell, VP of consumer product management, said the goal is to make connectivity simple across the devices people use most.

The operator positioned the launch as part of a broader effort to simplify connectivity and provide flexible options without long-term commitments.

Model compatibility ranges from seventh-generation iPads to the latest Pro and Air models with M3 chips.

AT&T ultimately plans to expand the service to other 5G-enabled devices including Android tablets, smartwatches, laptops and drones.

Multi-day options are also planned.

AT&T noted data rates may be temporarily reduced during periods of network congestion.

The post AT&T launches $3 daily iPad data pass appeared first on Mobile World Live.

China stepped-up preparations for 6G, launching a programme comprising regional pilots to identify potential use cases and make other advances towards eventual commercial rollout, state-owned news agency Xinhua reported.

The “joint ministry-provincial collaborative pilot program”, announced by China’s Ministry of Industry and Information Technology (MIIT), involves general research into 6G and how it fits into the wider mix of current and emerging technologies.

Among the focuses is exploration of potential applications for the network tailored to specific demand in the regions where the trials are set to take place, including smart maritime operations and use cases for manufacturing.

MIIT’s general aims for the project include the integration of communications infrastructure with AI, satellite and wireless sensing, alongside R&D into new base station technology and other network kit.

The news agency noted the move was intended to support eventual commercial rollout of the technology, which is expected around 2030.

MIIT’s announcement comes around a month after it gave the green light for regional trials in the 6GHz band.

The post China seeks 6G use cases in latest trials appeared first on Mobile World Live.