AMD shows first 7nm automotive FPGA and embedded processor
AMD is launching its first automotive parts on 7nm process technology with a focus on edge AI and cockpit
The 7nm AMD Versal AI Edge XA adaptive SoC includes hardware AI accelerators alongside an FPGA fabric and dual ARM processor cores for sensors for advanced driver safety and autonomous driving.
The automotive qualified 7nm Ryzen Embedded V2000A Series processor is aimed at infotainment and cockpit applications as part of the move to software defined vehicle architectures.
Versal AI Edge XA adaptive SoCs add a dedicated AI Engine for forward cameras, in-cabin monitoring, LiDAR, 4D radar, surround-view, automated parking and autonomous driving. The SoCs are also the first AMD 7nm device to be auto-qualified, bringing hardened IP and added security to automotive applications.
- Xilinx targets Versal FPGAs at edge AI
- Versatile computing platform could solve auto industry issues
- AMD introduces largest FPGA-based adaptive SoC
The SoCs can perform AI inference on large amounts of incoming data, and can also be used in edge sensors, such as LiDARs, radars and cameras, or in a centralized domain controller.
The AI Engines are capable of handling different types of AI models such as classification and feature tracking alongside a fabric of 20k LUTs to 521k LUTs, and from 5 TOPs to 171 TOPs. This allows designers to easily port designs with the same tools, ecosystem and safety certifications. The first devices will be released early 2024, with further releases planned later in the year.
“In the future, automakers will leverage autonomous vehicle applications to shape their brand identities. With these applications relying heavily on artificial intelligence, automakers need compute platforms that deliver powerful and efficient AI compute,” said James Hodgson, research director at ABI Research. “The number of highly automated vehicles* shipping each year is set to grow at a CAGR of 41% between 2024 and 2030, signaling a healthy growth opportunity for suppliers of heterogenous SoCs with powerful and efficient AI compute, including the AMD Versal AI Edge XA.”
- AMD launches 7nm embedded processors
- First 5nm embedded processors with Radeon graphics
- AMD and ECARX collaborate on in-vehicle computing
The Ryzen Embedded V2000A Series processors are also built on 7nm and are the first x86 auto-qualified processors to offer the same PC-like experience consumers have come to expect from home entertainment which they can now enjoy in-vehicle, on the go, says AMD.
“Our expanding and highly diversified AMD automotive portfolio presents a significant opportunity to serve this high-growth market while also underscoring the tremendous synergy of our combined automotive teams since the acquisition of Xilinx almost two years ago,” said Salil Raje, senior vice president and general manager, Adaptive and Embedded Computing Group, AMD.
This combines the 7nm Zen 2 cores and high-performance AMD Radeon Vega 7 graphics engine for high-definition graphics alongside enhanced security features and automotive software enablement through hypervisors in addition to support for Automotive Grade Linux and Android Automotive.
“Since introducing the ECARX Makalu digital cockpit using AMD Ryzen Embedded processors, they have proven to be key for powering the ECARX automotive solutions that require advanced compute power with visual graphic rendering capabilities,” said Peter Cirino, Chief Operating Officer, ECARX. “With the Ryzen Embedded V2000A Series processor, ECARX looks forward to expanding the capabilities of our next-generation digital cockpit solutions for software-defined vehicles in 2024 and beyond.”
Demonstrations of both technologies at CES 2024 include In-Vehicle Experience, AI Optimized Real-time Multi-sensor Object Detection, Automated Parking, Display Expansion, LiDAR, 3D Surround View and Driver Monitoring. Participating automotive ecosystem partners include: BlackBerry, Cognata, ECARX, Hesai, Luxoft, QNX, QT, Robosense, SEYOND, Tanway, Visteon and XYLON
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
