Specifications
SoC
Mobilint REGULUS (NPU included)
CPU
Quad-core Arm Cortex-A53 + Cortex-M0+
AI Performance
10 TOPS (INT8)
Thermal Design Power
8 W
Memory
8 GB LPDDR4X
Storage
32 GB eMMC v5.1 + 256 Mbit SPI Flash + microSD
Module
MLM-1 SoM, 55 x 50 mm — mounted on the carrier board
Carrier Board Size
100 x 110 mm
Power
DC +12 V to carrier; 5 V input to the SoM
Boot Mode
Standalone — boots Linux and SDK qb modules without a host PC
Camera Input
Connect a camera and run inference directly on the kit
Software
Linux 5.15 with Mobilint SDK qb modules preloaded
Supported Frameworks
Keras, TensorFlow, TensorFlow Lite, PyTorch, ONNX
Design Path
Develop on the EVK carrier, then design a custom carrier for the target product
Model Coverage
490+ open and closed-source models validated on the NPU architecture
Price
Quote upon request
Availability
Available
Overview
The REGULUS AI EVK is a complete standalone AI system that operates independently of a host computer. Connect a camera or other data interface, apply 12 V power, and the board boots Linux with Mobilint's SDK qb modules preloaded — a much shorter route to a working prototype than a host-tethered accelerator card.
The kit consists of two boards: the MLM-1 SoM with the REGULUS SoC mounted on top, and a 100 x 110 mm carrier board carrying the I/O ports. The SoC itself integrates the NPU, a quad-core Arm Cortex-A53 cluster with a Cortex-M0+, to deliver 10 TOPS of INT8 inference at an 8 W thermal design power, with 8 GB of LPDDR4X, 32 GB of eMMC v5.1, 256 Mbit SPI flash and microSD storage.
Mobilint designed the EVK as a development bridge: teams validate their application on the EVK carrier board, then design a custom carrier for the production product. The compute module, software stack and compiled models carry over with no changes, so the validated model set is not invalidated by the hardware transition.
QS Compute supplies the REGULUS AI EVK alongside the MLM-1 SoM, MLX-R1 mini AI box and MLD-R1 USB accelerator, giving teams the full REGULUS development-to-deployment chain.
Key Benefits
Standalone operation — no host PC needed to evaluate the SoC. 10 TOPS at 8 W with NPU, CPU, ISP and codec on one die. SDK qb preloaded with Linux 5.15. Direct migration from EVK carrier to a custom production carrier. Validated across 490+ models.
Applications
Edge AI product prototyping; machine-vision and camera application development; robotics and AGV perception validation; industrial IoT analytics pilots; LLM and multimodal inference evaluation on low-power silicon; academic and research platforms for efficient inference.
Request a Quote — MOBILINT REGULUS AI EVK — 10 TOPS STANDALONE EDGE AI DEVELOPMENT KIT
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →