Intel is developing a dedicated processor for the vehicle family - A3900. This family of CPUs will make it possible to build a fully software-defined cockpit solution that integrates functions such as in-vehicle infotainment systems, digital dashboards and advanced assistive driving systems into cost-effective SoC system-on-a-chip.
Along with the new generation Atom processor E3900 for the Internet of Things (IoT), Intel has released the new car processor. The A3900 will allow carmakers to build the capabilities of a higher-level real-time decision-making system for a new generation of vehicles. The chip is currently in beta, will be officially listed in the first quarter of 2017.
Intel Atom's new automotive-specific processor offers excellent memory access speed and bandwidth (up to LPDDR4 2400 and 38.4 GB / s, respectively) in a small, low-power package. For media applications it provides high-speed graphics and high-definition video capabilities, support 4K video at 60Hz. At the same time Intel TCC (Time Coordinated Computing) technology can make computing-intensive and real-time requirements of the video requirements for intelligent network of vehicles. The TCC coordinates ports and networked devices to achieve optimal decision-making at the lowest latency.
The A3900 integrates quad-core, energy-efficient CPUs, a robust GPU and dedicated audio, video and image processor, and the ability to handle multisensor-related tasks. The FCBGA package and Intel 14nm technology, the chip has a new graphics processing engine, compared to the previous Atom products provide doubled graphics processing power, but also greatly improve the video processing capabilities.
Intel VT for Directed I / O, which isolates device allocation and access, ensures that in-car entertainment information devices (IVIs) and dashboards are secure when accessing the processor at different levels of priority. nud4700snt1g
A3900 as a vehicle computing module also has a centralized management of energy consumption and memory functions. The application module is certified to the integrated circuit pressure test in the Automotive Electronics Council AEC-Q100 standard and has a seven-year industrial lifetime at a nominal temperature of -40 ° C to 110 ° C. The module is pre-validated on Intel automotive software tools and multiple operating systems, so suppliers can simplify the design of PCBs based on it.
The processor's advanced hardware-level protection provides robustness, including an integrated security engine and a dedicated security coprocessor that dynamically adapts security levels for mission-critical tasks. It also provides secure bootstrap and fast encryption execution with the help of the Intel Advanced Encryption Standard New Instruction Set (Intel AES-NI).
To accelerate and simplify application development, Intel and its partners offer a comprehensive set of development tools and third-party operating system support, including:
Intel C + + compiler, VTune Amplifier, graphics performance analyzer (Graphics Performance Analyzer)
The reference stack application includes an IVI middleware and a vehicle boot loader boot loader
Operating system support, including Linux, Android Auto, Green Hill, QNX, and Wind River Helix Cockpit
QNX, Green Hill, Mentor Graphics multi-operating system super management procedures
Based on Intel CPU and GPU performance debugging tools and a full set of hardware development vehicles
In addition, Atom's E3900 series will make edge computing (edge computing) and fog computing (fog computing) more intelligent. It will cause more calculations to occur around the data sensor and reduce the computational pressure on the data center side. Fog computing, also known as fog networking, is a de-centralized computing architecture that enables computing resources and application services to be more logically distributed between data sources and clouds. tps54620rgyt
For example, taking into account data from traffic cameras and sensors, there is now a clear downward trend in sending data to server-side analysis, which is related to the delay caused by video compression and data transfer, and the processing capacity on the terminal side These shortcomings can be overcome.
In the automotive industry, the edge computing power will also give the software definition of the cockpit to bring new changes. Independent system computing capabilities that drive digitized dashboards, navigation and advanced assistive driving functions will be a trend. These capabilities are required for the rear view sensor, the parking imaging system, and the side impact avoidance function.
- baitron's blog
- Log in or register to post comments