Compute Solutions
Edge AI Computing Solutions which can Simplify, Accelerate, and Reduce the Cost of Embedded Systems Development
Virtium’s new SOM product family combines the proven innovation of Embedded Artists with Virtium’s global manufacturing, quality, and support infrastructure. Backed by more than 20 years of Embedded Artists’ expertise in SOM design, Edge AI computing, and connectivity, and strengthened by Virtium’s leadership in memory, storage, and industrial-grade solutions, these products deliver reliable, high-performance platforms for Edge AI and embedded applications. Customers can confidently deploy Virtium SOM solutions knowing they are supported by robust engineering, long-term lifecycle management, and high-volume manufacturing in the USA and Vietnam.
RZ/G3E DX-M1 SOM
Experience a new level of intelligent performance with our new System-on-Module, seamlessly integrating the versatile power of the Renesas RZ/G3E application processor with the groundbreaking DEEPX DX-M1 AI Booster. This powerful combination delivers exceptional processing capabilities and unprecedented AI performance!
Options for implementing AI inference at the edge
The RZ/G3E SOM gives developers two options for edge AI processing. The RZ/G3E processor includes an integrated Arm Ethos™-U55 neural processing unit (NPU) core rated for 0.5 TOPS of inference performance, suitable for entry-level vision AI, voice AI and sensor data-based AI applications.
Virtium Embedded Artists also supplies the SOM in a version with a DeepX DX-M1 AI accelerator which offers an impressive 25 TOPS of AI performance, for highly sophisticated inference applications.
Key Features:
- RZ/G3E: 4x Cortex-A55 @1.8 GHz & Cortex-M33 @200 MHz delivering 21.6K DMIPS
- Dual display capable via LVDS, MIPI-DSI or parallel
- H.264/265 Enc./Dec., Mali-G52 GPU@30GFLOPS
- USB3.2, USB2, Dual Gigabit Ethernet
- DEEPX DX-M1 AI Booster with 25 TOPS at 5 Watt!
VMX-004:
AI ACCELERATOR Module
The module conforms to the PCIe M.2 2280 standard, allowing seamless integration into most ARM and x86 carrier boards and motherboards for fast, hassle-free deployment. Designed and assembled in Southern California, the M.2 module enables OEMs and solution developers to deploy high-performance Vision AI across a wide range of markets.
Key Features:
- Four MemryX MX3 AI processors
- Up to 24 TOPS AI Computing performance
- Industrial-temperature operation: -40°C to +85°C
- Form factor M.2 2280
- Interface PCIe Gen3 2×2 lanes
- Typical Power: 3 – 10 Watts
- OS Support: Ubuntu, Windows and Android
- Frameworks: ONNX, PyTorch, TensorFlow, Keras , TensorFlow Lite
All MPU SOMs
Virtium – Embedded Artists’ specialize in developing System-on-Modules (SOMs) that can be easily integrated into your product – with many benefits:
- You can focus on your core business!
- You shorten your time to market!
- You lower your risk and cost by using our professionally designed and proven modules.
A uCOM/COM/SOM board integrates all core components around a specific NXP or Renesas application processor. We take care of development, production and life cycle management – you focus on your application!
When integrating a uCOM/COM/SOM board into a design our Developer’s Kits gives you a flying start. In many cases, our pre-designed carrier boards can be directly integrated in your product or serve as reference implementations.
All MCU SOMs
Embedded Artists’ specialize in developing microcontroller (MCU) boards that can be easily integrated into your product – with many benefits:
- You can focus on your core business!
- You shorten your time to market!
- You lower your risk and cost by using our professionally designed and proven modules.
A MCU board integrates all core components around a specific NXP microcontroller. We take care of development, production and life cycle management – you focus on your application!
When integrating a MCU board into a design our Developer’s Kits gives you a flying start. In many cases, our pre-designed carrier boards can be directly integrated in your product or serve as reference implementations.
RZ/G3E DX-M1 SOM
RZ/G3E DX-M1 SOM seamlessly integrating the versatile power of the Renesas RZ/G3E application processor with the groundbreaking DEEPX DX-M1 AI Booster. This powerful combination delivers exceptional processing capabilities and unprecedented AI performance!
M.2 VMX-004
M.2 module powered by MemryX MX3 delivers 24 TOPS of high-efficiency AI performance for vision application. AI models run entirely from on-chip SRAM, eliminating external DRAM to reduce latency and improve performance with lower power.
All MPU SOMs
A uCOM/COM/SOM board integrates all core components around a specific NXP or Renesas application processor. We take care of development, production and life cycle management – you focus on your application!
All MCU SOMs
A MCU board integrates all core components around a specific NXP microcontroller. We take care of development, production and life cycle management – you focus on your application!
Virtium's Product Portfolio
