SolidRun BEDROCK R7000 EDGE AI AMD Ryzen 7 7040 Series Edge AI Box PCs with Hailo-8 Accelerator

Camera-based detection at the edge usually stalls on one of two things: not enough inference throughput, or a computer that cannot survive the cabinet it is fitted in. SolidRun built the BEDROCK R7000 EDGE AI series to address both, combining an eight-core AMD Ryzen 7 7040 Series processor with Hailo-8 AI acceleration modules in a fanless, all-aluminium enclosure. SolidRun aims these edge AI box PCs at vision-based situational awareness in harsh environments, from Industry 4.0 lines to autonomous guided vehicles.

Each BEDROCK R7000 EDGE AI unit shares the same SoM, NIO R7000 Basic and SX 4M2 boards. The Ryzen 7 7840U or 7840HS gives eight Zen 4 cores and 16 threads at up to 5.1 GHz, Radeon 780M graphics and a 10 TOPS NPU, and 20 native PCIe Gen4 lanes let up to three Hailo-8 modules run alongside NVMe Gen4 x4 storage, for up to 78 TOPS of combined inference. Two M.2 key-M sockets take a Hailo-8 or an NVMe SSD each, and the key-E socket takes a third Hailo-8 or Wi-Fi 6E. Two DDR5 SO-DIMM sockets take up to 64 GB with ECC, beside two Intel I226 2.5GbE ports, four USB 3.2 ports and four display outputs.

The BEDROCK R7000 EDGE AI 30W is the compact finned build at 45 mm wide and 0.9 litres, cooling up to 30 W by convection on a crowded DIN rail. The BEDROCK R7000 EDGE AI 60W moves to a 73 mm, 1.5 litre enclosure that dissipates up to 60 W, so the Ryzen 7 and the AI modules can work under sustained load. The BEDROCK R7000 EDGE AI TILE is 29 mm thick with flat walls for conduction cooling into a cold plate, for vehicles and sealed enclosures. Hailo options and I/O match across the three.

Power enters through a screw-locking Phoenix terminal at 12 to 60 V DC, with redundant BIOS flash, a watchdog timer and TPM 2.0 for unattended sites. An optional Quectel 4G or 5G modem with two nano-SIM trays covers remote cameras. Rated up to -40 to 85 °C and mounting on a DIN rail, wall, VESA bracket or table top, these edge AI box PCs suit robotics, transport, smart city, retail and agriculture vision on Windows or Linux.