SAPPHIRE unveils EDGE+ RAVE2 MAX for automotive Physical AI
SAPPHIRE Technology has launched the EDGE+ RAVE2 MAX, the top-of-range member of its RAVE2 embedded compute family, aimed at automotive AI, autonomous mobility, machine vision and industrial systems. The platform ships in two configurations: a Flex-ATX board (229 x 191mm) and a complete embedded system, giving OEM integrators and solution developers flexibility over deployment approach.
The product was formally showcased at AMD's Advancing AI 2026 event in San Francisco and is available from today, 28 July 2026.
The platform
RAVE2 MAX is built on AMD's Embedded+ architecture, combining an AMD Ryzen Embedded X100 series processor with an AMD Versal AI Edge Series Gen2 VE3558 adaptive SoC. That pairing divides workloads by character: the Ryzen X100 handles x86 general-purpose compute, graphics acceleration via integrated AMD Radeon graphics, and human-machine interface rendering, while the Versal device contributes its FPGA fabric, programmable I/O and AI Engines for deterministic, low-latency inference and control loops.
On paper the specification is designed for high-channel sensor rigs. The I/O board provides 12 GMSL2 camera ports, enabling full surround-camera perception, alongside dual CAN Bus interfaces for in-vehicle networking. Host memory tops out at 128GB LPDDR5x across eight modules; the Versal SoC carries an additional 16GB LPDDR5x. Connectivity includes dual 10GbE, USB4 Type-C, OCuLink PCIe Gen4 x4 expansion and five or more mini-DisplayPort outputs. Security is handled by an Infineon OPTIGA TPM 2.0 module. The board runs from a 12VDC supply and supports embedded Linux alongside unspecified commercial OS offerings.
Paul Smith, Senior Director of SAPPHIRE Embedded Solutions, said the platform is intended to let customers "replace complex multi-board designs with a single validated platform and bring sensor-rich automotive AI systems to market faster." Simon George, Director of Embedded Systems, Physical AI at AMD, characterised the RAVE2 family as demonstrating how the Embedded+ silicon stack can "consolidate complex workloads and accelerate the deployment of automotive, mobility and industrial AI systems."
Market context
Physical AI at the edge is a crowded and fast-moving segment. SAPPHIRE competes alongside companies including Advantech, Congatec, and a range of system-on-module vendors building on NVIDIA Jetson and Qualcomm automotive-grade silicon. NVIDIA's Drive platform in particular has established significant traction with ADAS and robotaxi programmes, meaning AMD-ecosystem partners such as SAPPHIRE must make a credible case on price-performance, software support and supply-chain stability rather than name recognition alone.
The Embedded+ architecture is AMD's answer to that challenge: a co-validated bundle of Ryzen and Versal silicon, reference designs and software enablement intended to reduce the bring-up burden for OEMs. Extending the family upward with the RAVE2 MAX, and maintaining a common software foundation across RAVE2 Core, Pro and MAX, is a sound strategy for preserving software investment as customers scale channel counts or inference workloads.
Regulatory tailwinds are building for automotive AI hardware. UN Regulation 157 on automated lane-keeping systems is expanding in scope across major markets, and the EU's forthcoming General Safety Regulation requirements for ADAS on new vehicles will create structured demand for production-validated compute platforms through the late 2020s. Functional-safety certification (ISO 26262, SOTIF) and cybersecurity compliance (UN R155) are increasingly prerequisites rather than differentiators for OEM procurement conversations, and SAPPHIRE has not yet stated what certification status the RAVE2 MAX holds against those standards.
Pricing and formal availability timelines beyond the launch announcement have not been disclosed. The company says customers can obtain further information through the SAPPHIRE commercial embedded product page.