Texas Instruments Incorporated (TI), a U.S. semiconductor corporations, unveiled automotive chips and development solutions that will accelerate the implementation of "Autonomous Driving level 3" (L3). Under the Society of Automotive Engineers (SAE) standards, L3 is a level of "conditional Autonomous Driving," in which the driver intervenes only in specific situations while the vehicle drives itself in most sections.
TI on Jan. 6 (local time) will unveil three products at CES 2026, the world's largest consumer electronics and IT show in Las Vegas, that can meet L3 requirements. ▲ TDA5 high-performance computing system-on-chip (SoC) ▲ 4D imaging radar transceiver (product name AWR2188) ▲ Ethernet physical layer (in-vehicle communications that transmit data generated by Autonomous Driving vehicle sensors, product name DP83TD555J-Q1 10BASE-T1S). The company said these can raise customers' technology levels.
Before CES 2026 opens, TI held a pre-briefing for media in the Asia-Pacific (APEC) region, including Korea, Japan, China, Taiwan, the Philippines and Malaysia, and said, "The TDA5 product line, which combines power-efficient, safe processing technology with Edge AI (technology that performs AI computation directly on devices where data is generated), along with the new radar transceiver and in-vehicle communications device, supports L3 Autonomous Driving."
Mark Ng, TI's head of automotive systems, said, "The auto industry is moving toward less driver intervention," adding, "By applying TI semiconductor technology that enables safe Autonomous Driving, we can innovate across sensing, communications and decision-making."
◇ "Developing AI chips that can achieve L3… best-in-class power efficiency"
The TDA5 chip, which serves as the "brain" of an Autonomous Driving vehicle, is the core of TI's new solution. By applying this chip, it is possible to build an "Edge AI" environment that can expand operations per second from 10 trillion (10 TOPS; 1 TOPS is 1 trillion operations per second) to as high as 1,200 trillion (1,200 TOPS). The company said that this enables vehicles to quickly analyze and respond to data even when encountering diverse road environments, thereby making L3 possible.
Power efficiency is also a strength. The chip supports 24 TOPS of computation per watt (W). Roland Schupfli, head (vice president) of TI's processor product institutional sector, said, "For EVs, how far they can go on a single charge is a key metric, so you need chips that use less power while delivering high performance," adding, "TDA5 offers the industry's best power efficiency."
To realize the low-power, high-performance TDA5 chip, TI integrated its Neural Processing Unit (NPU) product C7. Vice President Schupfli said, "We achieved up to 12 times higher AI computing performance than the previous generation while keeping power consumption at a similar level," adding, "We built functions to enable broad integration of safety, security and computing applications."
TI said the TDA5 supports a chiplet (a technology that splits high-performance semiconductor functions into multiple parts and then combines them) design, enabling customized applications for customers. To that end, the company built the TDA5 based on UCIe (an open standard interface that efficiently consolidations and communicates among chiplets within a single package).
TI also provides a virtual development kit in collaboration with Synopsys to improve customers' development efficiency. Through development tools that include a digital twin (a concept that precisely implements the real world in a digital space for simulations), customers can test how TDA5 can be applied to their products without a vehicle. The company described it as a function "that helps shorten the time to market for software-defined vehicles (SDV) by up to 12 months."
◇ "High-definition radar that works even in bad weather… Ethernet capable of nanosecond synchronization"
The 4D imaging radar transceiver launched alongside TDA5 provides eight transmitters and eight receivers (8x8) via a single LOP (antenna-in-package technology with direct package consolidation). Without adding separate devices to the vehicle, it can implement a high-resolution radar system, simplifying vehicle design. Keegan Garcia, TI's high-performance radar product line manager, said, "Radar has an advantage for Autonomous Driving compared with other sensing systems and works without issues even in bad weather," adding, "TI's radar solution recognizes objects with high accuracy even at 350 meters." Garcia added, "Because the solution supports eight transmitters and eight receivers, it reduces the number of devices needed even if customers expand to more channels," and "It supports both satellite and edge architectures, making it suitable from compact cars to premium vehicles."
For this radar transceiver, TI applied an analog-to-digital converter (a device that converts physical signals into data) and a radar chirp signal slope engine (a technology that calculates the rate of change of the radar's transmitted signal to measure distance more accurately), improving performance by up to 30% compared with existing products.
The Ethernet solution that carries in-vehicle signals has also advanced. The Ethernet serial peripheral interface (SPI), a communication method for exchanging data between chips, integrates a media access controller (MAC), which manages data transmission order and collisions. The company said this enables synchronization at the nanosecond level (one-billionth of a second), enhancing vehicle safety.
TI has already sent prototypes of the newly unveiled Autonomous Driving vehicle solutions to some automakers to evaluate performance. At CES 2026, the company is operating a booth that lets attendees experience a vehicle environment with the new solutions applied.
Director Mark Ng said, "Based on roughly 20 years of TI's automotive processing experience, we have greatly expanded our business portfolio with the new solutions," adding, "We will support not only automakers but also parts suppliers in implementing advanced AI models."