AMD launches Embedded+ architecture

Compared to mobile, desktop, and server processors, the general interest in AMD’s embedded products may not be as pronounced. However, AMD’s embedded processors are of paramount importance to edge devices, including sectors such as industrial, automotive, medical, digital gaming consoles, and compact client systems. Today, AMD unveiled its latest embedded architecture, Embedded+, which amalgamates the Ryzen embedded processors based on the Zen+ architecture with the Versal adaptive SoC onto a single PCB.

The Embedded+ architecture synergizes the capabilities of Ryzen embedded processors with the Versal AI Edge adaptive SoC within a unified package. AMD targets critical domains requiring robust computational power and energy efficiency. This collaboration enables Embedded+ to manage real-time AI inference and complex sensor data, indispensable for applications in dynamic and demanding environments.

Integrating Ryzen embedded processors and Versal SoC, this architecture especially benefits industries necessitating low latency between hardware and software, including autonomous vehicles, diagnostic equipment in healthcare, and precision machinery in industrial automation. Moreover, AMD Embedded+ supports a diverse workload across different processor types, including x86 and ARM, as well as AI engines and FPGA structures, offering flexibility and scalability for embedded computing solutions within the industry.

The AMD Embedded+ platform ensures extensive compatibility with various sensor types and their corresponding interfaces. It facilitates direct connections with standard peripherals and industrial sensors through Ethernet, USB, and HDMI/DP interfaces. The architecture’s AMD Ryzen embedded processors can process inputs from traditional image sensors, such as RGB, monochrome, and even advanced neuromorphic types, while supporting industry-standard image sensor interfaces like MIPI and LVDS.

The AMD Versal AI Edge adaptive SoC on the Embedded+ motherboard further augments its capabilities, providing adaptive I/O options for real-time sensor inputs and industrial networks. This includes interfaces with LIDAR, radar, and other sophisticated sensors required by modern embedded systems in industrial, medical, and automotive fields. The platform supports a variety of product-grade sensor interfaces, such as GMSL and Ethernet-based vision protocols, indicating its meticulous design for integration into complex sensor-driven systems.

AMD also announced the launch of new pre-integrated solutions available for ODM manufacturers from today. The Sapphire VPR-4616-MB platform, featuring an ITX form factor, employs AMD Versal AI Edge 2302 SoC and AMD Ryzen Embedded R2314 processor, based on the Zen+ architecture, with four cores and threads, and is equipped with six Vega CU units. The motherboard is equipped with two DDR4 SO-DIMM memory slots, supporting up to 64GB of memory, a PCIe 3.0 x4 M.2 slot, and a SATA port. In terms of networking, it offers a 2.5 Gb NIC and an M.2 Key E slot for expanding WiFi capabilities and supports the Linux-based Ubuntu 22.04 operating system.

Additionally, a series of expansion boards have been released, significantly extending support for the Embedded+ architecture. Notably, the Octo GMSL camera I/O board can connect with multiple cameras simultaneously, making it suitable for high-bandwidth vision-based systems used in advanced driver-assistance systems and automated monitoring systems. These systems often require the integration of numerous image inputs for real-time processing and analysis, and the Octo GMSL board is designed to meet this need.

Furthermore, there is a dual Gigabit Ethernet I/O board for environments requiring high-speed network communication. The dual 10 Gb SFP+ board, with 16 GPIOs, meets higher bandwidth demands, providing ample data transfer rates for real-time video streaming and large-scale sensor data aggregation. These expansion options broaden the Embedded+ architecture’s capabilities in edge and industrial scenarios.