Home » AMD Versal Gen2, AI inference and end-to-end post-processing

AMD Versal Gen2, AI inference and end-to-end post-processing

by admin
AMD Versal Gen2, AI inference and end-to-end post-processing

AMD expands adaptive SoC portfolio with Capital Gen 2targeting 3x TOPs-per-watt and offering preprocessing, AI inference, and postprocessing.

The architecture also offers up to 10 times more modular CPU-based computing than the first generation.
Subaru is among the first customers to announce plans to use AMD Versal AI Edge Series Gen 2 to power its next-generation “EyeSight” ADAS system.

These first models in the Versal Series Gen 2 program feature powerful new AI engines that will deliver up to 3x more TOPs-per-watt than first-generation Versal AI Edge Series devices, while new integrated high-performance Arm CPUs will deliver scalar calculation up to 10 times higher than first generation Versal AI Edge and Prime Series devices.

Balancing performance, power and area, along with advanced functional safety, Versal Series Gen 2 devices deliver new capabilities and features that enable you to design high-performance, optimized products for the automotive, aerospace and defense, industrial, vision, healthcare, broadcast and professional AV.

Subaru’s next-generation ADAS vision system

Subaru Corporation has selected Versal AI Edge Series Gen 2 devices for its next-generation Advanced Driver-Assistance System (ADAS) vision system, known as EyeSight. This device is integrated into some Subaru car models to enable advanced safety features, including adaptive cruise control, lane keeping assist and pre-collision braking. Subaru uses AMD adaptive SoC technology in current EyeSight-equipped vehicles.

Versal AI Edge Series Gen 2

To meet the complex computing needs of real-world systems, AMD Versal AI Edge Series Gen 2 devices integrate an optimal mix of processors for all three stages of AI-based embedded system acceleration:

Pre-processing – FPGA programmable logic for real-time multiprocessing with unprecedented flexibility that allows you to connect to a wide range of sensors and implement high-throughput, low-latency data processing pipelines.
AI inference – A set of vector processors in the form of next-generation AI engines to achieve efficient AI inference.
Post-processing – Arm CPU cores capable of providing the post-processing power needed to make complex decisions and control the operations of safety-critical applications.

See also  30 series inventory backlog is serious? N card shipments plunged 25.7% in the second quarter – Hardware – cnBeta.COM

This single-chip intelligence can make the creation of multi-chip computing solutions unnecessary, resulting in smaller, more efficient embedded AI systems, resulting in shorter time-to-market potential.

Versal Prime Series Gen 2

AMD Versal Prime Series Gen 2 delivers end-to-end acceleration for traditional non-AI-based embedded systems by combining programmable sensor processing technology with high-performance embedded Arm CPUs. Designed to deliver up to 10x more scalar processing than the first generation, these devices can efficiently handle sensor processing and complex scalar workloads.

Featuring new hard IP for high-speed video processing, including multi-channel workflows up to 8K, Versal Prime Gen 2 devices are ideal for ultra-high definition (UHD) video recording and streaming applications, as well as industrial PCs and flight computers.

Large and scalable portfolio

Versal AI Edge Series Gen 2 and Versal Prime Series Gen 2 products offer scalability from edge sensors to centralized compute for AI-driven systems. They feature a range of devices with an increasing number of AI and adaptive compute to allow customers to select performance, power and compute area to efficiently achieve performance goals and application security.

Optimized design cycles

The tools and the library of AMD Vivado Design Suite help increase productivity and simplify design cycles for system developers hardware embedded, offering fast compilation times and better quality of results. For embedded software developers, the AMD Vitis Unified Software Platform enables the development of AI design, signal processing, and embedded software at users’ preferred levels of abstraction, without requiring any experience with FPGA.

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More

Privacy & Cookies Policy