- Artificial Intelligence research venture between UTS and Chinese government
- Built-in power supply for temperature controller
- Panasonic opens lithium-ion battery factory in China
- Resistors for power management applications
Texas Instruments has unveiled an automotive System-on-Chip (SoC) family, the TDA2x, incorporating a Vision AccelerationPac with features to help customers create advanced driver assistance systems (ADAS) that help reduce the number of collisions on the road and enable autonomous driving experiences.
Unlike any other processors in the industry, TI's integrated TDA2x device family combines a mix of performance, vision analytics, video, graphics and general purpose processing cores in a low power envelope, enabling a range of ADAS applications including front camera, surround view and sensor fusion.
Furthermore, TI's Vision AccelerationPac complements the TMS320C66x DSP generation cores, and enables more ADAS algorithms to run simultaneously.
With the TDA2x family, TI's customers will be able to achieve embedded automotive vision by enabling the range of ADAS applications on a common architecture, allowing algorithm investments to be shared across multiple systems.
The TDA2x enables front camera applications including high beam assist, lane keep assist, advanced cruise control, traffic signal recognition, pedestrian / object detection, and collision avoidance to run simultaneously.
Additionally, the TDA2x enables park assist applications including intelligent 2D and 3D surround view and rear collision warning and can also run a pedestrian/object detection algorithm developed for the front camera. TI's TDA2x also serves as a central processor for fused radar and camera sensor data, allowing more robust ADAS decision making.
The TDA2x SoC is based on a heterogeneous, scalable architecture that includes TI's fixed-and floating-point C66x DSP cores, a fully programmable Vision AccelerationPac, ARM Cortex-A15 MPCore processors and dual Cortex-M4 cores along with video and graphics cores and a host of peripherals.
The Vision AccelerationPac is designed to run low to mid-level vision processing functions, offloading the DSP and ARM cores. The result is optimal performance at the lowest power footprint.
TI's programmable Vision AccelerationPac delivers more than 8x compute performance at the same power budget for advanced vision analytics in a more cost-effective footprint.
Built on TI's knowledge of algorithms and designed to meet the needs of the ADAS market, the Vision AccelerationPac contains one or more multiple Embedded Vision Engines (EVEs), to offer a purpose built ADAS accelerator.
Comprised of an optimised vector coprocessor and a 32-bit programmable RISC core, these engines allow for flexible development by automotive Tier 1s and OEMs. Additionally, the engines are capable of running simultaneous ADAS algorithms faster, and with greater power-efficiency than ever before.
TI offers royalty-free software libraries and framework to provide developers an optimised collection of vision and imaging functions to take advantage of the TDA2x and Vision AccelerationPac's architecture while reducing development efforts and time to market.