Texas Instruments and Qualcomm doubled down on new automotive semiconductors and related hardware on the eve of CES 2026, as executives from both companies described the future of the automotive vertical as vibrant despite recent churn.
Meanwhile, edge semiconductor maker Ambarella introduced prior to CES the CV7 edge AI vision system-on-chip, which the company described as ideal for automotive designs, in addition to consumer products, industrial automation, security cameras and robotics. All of these new products from the three companies support sensors of all types in use in automotive applications, including lidar, radar and vision sensors.
The focus on automotive at CES may be dampened as several sensors and sensors-related manufacturers recently told Fierce that they see stagnation in the auto sector due to various factors such as the congressional decision to drop the federal EV tax credit in the One Big Beautiful Bill. But executives for both TI and Qualcomm told Fierce they are planning now for 2026 through 2030 amid continued industry interest in level 3 autonomous vehicles, where the vehicle is considered the driver, and even higher levels of autonomy.
“I won’t disavow that the auto industry is under massive churn,” Qualcomm’s Mark Granger, vice president of product management for Snapdragon cockpit, said in response to a question from Fierce in a briefing with reporters. “We have a fully scalable platform…and whether it’s an EV or an ICE vehicle, there’s not a huge difference and the transition to EV is still a huge opportunity. There’s a hotbed of opportunity for Qualcomm products…People have woken up in the last year, understanding they really need generative AI in vehicles in the next couple of years. I don’t see any stagnation in our traction.”
Anshuman Saxena, general manager of ADAS and robotics at Qualcomm, said the company is seeing continuous growth in Snapdragon. “For what it’s worth it’s the opposite of stagnation…a very strong portfolio.” Qualcomm’s Snapdragon Digital Chassis technology already powers 400 million vehicles, the company noted.
(At CES, Qualcomm is also highlighting its new Dragonwing Q-7790 and Q-8750 processors for on-device AI, as part of its Augentix acquisition. The new Dragonwing IQ10 series is designed to power robotics as well. The biggest reveal at CES may well be its Snapdragon X2 Plus with its third gen Oryon CPU for up to 35% faster performance with PCs. )
Losses announced by automakers in 2025 including recently by Ford, and earlier by GM and Stellantis, are on the minds of device makers like Qualcomm and TI, but have not dampened interest in innovation, development and production.
“There’s a lot of negative news in the market every day, with recent heavy losses by automakers,” said Mark Ng, director of automotive systems at TI, in comments to Fierce. "But if you think of a future car, you’re on a journey and we're seeing more and more compute and displays that simply buck the overall narrative of a slowdown. We are very engaged and wouldn’t make these devices if we didn’t have that demand.”
L3 already requires hundreds of TOPS of AI compute, but full automation at L4 and higher will require over a thousand TOPS, TI estimates. TI introduced a TDA5 system on chip family that will deliver from 10 to 1200 TOPS of AI acceleration for vehicles. It provides up to 12 times the AI computing of prior generations, TI said in a pre-brief with reporters.
Ambarella’s CV7 can provide automakers with AI vision gateways and hubs for use in passive driver assistance systems and 360-degree surround view and video recording, as well as multiple processing of multiple video streams. It consumes 20% less power than its predecessors, partly due to Samsung 4nm process technology, the company said in a pre-briefing with Fierce.
TI also introduced an AWR2188 4D imaging radar transceiver for greater vehicle automomy in all weather conditions, useful in L3 vehicles and above. The change with 4D radar is the ability to add vertical angle measurement to a standard radar with measures of range, velocity and horizontal data, according to Yariv Raveh, vice president for radar at TI, in comments to Fierce.