<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1332818711964721&amp;ev=PageView&amp;noscript=1">

Meet the EP‑200Q – Built for Edge AI with AMC Edge.

Seamless. Always connected. Purpose-built for intelligent devices at the edge.

Explore the EP-200Q
Visit resource center

What is "Edge AI"?

Edge AI brings artificial intelligence closer to where data is created, at the edge.

Rather than sending data to the cloud, Edge AI devices process and analyze information locally. This enables real-time responsiveness, strengthens data privacy, and reduces dependency on constant network connectivity.

Selecting the right platform for your specific AI workload is essential. The right choice enhances power efficiency, reduces cloud storage and bandwidth demands, and ensures your system delivers optimal performance exactly where its needed most.

Featured Resources

Article
Why On-Device AI Is the Future of Inference
Learn more
Article
What kind of device is suitable for your on-device AI inference?
Read now
Article
Which architecture is the best for your edge AI device?
Read now

AMC Edge Solutions

EP-200Q

EP-200Q

AMC Edge SoM enabled by Qualcomm SoC QCS6490

  • Pre-validated QCS6490 subsystem
  • Always-on connectivity for edge AI
  • High-performance Wi-Fi 7

Q6_from_web_png

 

See product
EP-200Q-CB

EP-200Q-CB

Edge connector SoM with EP-200Q and Wi-Fi 7

  • Enabling vision AI on edge devices
  • Always-connected edge devices
  • Dedicated support for seamless integration

Q6_from_web_png

See product

Edge AI for Industries

Enabling Edge AI devices with robust Wi-Fi connectivity to deploy them everywhere.
AdobeStock_1492600537

Edge AI in Healthcare

By integrating intelligence at the edge, Edge AI optimizes healthcare operations, enriching patient care and streamlining the day-to-day experience of healthcare professionals.

Read article
AdobeStock_554220783

Edge AI for Automation

By bringing Edge AI-powered vision to the edge, Edge Machine Vision enables adaptive robots to make real-time decisions, driving greater productivity and efficiency to warehouse operations.
Read article
AdobeStock_1311226764

Edge AI for Safety

By processing data locally, Edge AI cameras enhance real-time safety and security across diverse settings, integrated with robust connectivity, they ensure continuous protection and situational awareness.
Read article
Edge AI resource center

Edge AI resource center

Your one-stop destination for edge AI knowledge: hardware insights, software frameworks, and deployment strategies.
Browse now

Partners

Silex partners with leading-edge technology companies to enable edge AI solutions.
Qualcomm Logo

Qualcomm

Qualcomm  is a global leader in innovative technologies for on-device AI solutions. Silex Technology is a certified Qualcomm Embedded Design Center (EDC), providing  Dragonwing SoC and wireless technology implementation support. Learn more about our partnership with Qualcomm. 

Frequently Asked Questions

What kind of embedded devices is the EP-200Q for?

The EP-200Q is suitable for embedded devices that integrate multiple cameras to perform vision AI inference locally.

The EP-200Q is powered by Qualcomm's QCS6490, a high-performance and power-efficient processor. Therefore, battery-powered mobile or portable embedded devices will benefit from the EP-200Q.

Silex's expertise in Wi-Fi and SX-PCEBE support will add robust wireless connectivity for the devices.

Here a few articles about possible use cases.

Edge vision AI in healthcare

Edge vision AI for automation

Edge vision AI for safety

How is the EP-200Q compared with other processors available on the market?

The EP-200Q's CPU performance is  comparable to higher-end of Intel Atom processors, NVIDIA Jetson Xavier NX, Orin Nano, or similar SoCs from other manufacturers.

The EP-200Q's NPU provides 12 TOPS of neural processing power, which is comparable to the Jetson Xavier NX or the Orin Nano 4GB when running dense models. It is also comparable to Hailo's 15-TOPS AI vision processor.

One of the EP-200Q's advantages is its power efficiency. When compared with the Jetson Orin Nano running in its 7W power mode, the EP-200Q can deliver better performance, although the Jetson Orin Nano's peak performance is still higher than that of the QCS6490.

What kind of AI inference can the EP-200Q handle?

The EP-200Q's hardware features and SDK are good fit for vision AI applications.

The EP-200Q's SDK is based on one of the followings:

  • Qualcomm Intelligent Multimedia SDK (QIMP)
  • Qualcomm Intelligent Robotics SDK (QIRP)

Both SDKs are targeted more toward  vision AI and/or sensing AI applications rather than Gen AI or agentic AI applications.

Applications requiring generative AI may be possible by using an SLM (Small Language Model) optimized for specific use cases, but no SLM/LLM has been validated with the EP-200Q.

Can AI models available today be used with the EP-200Q?

It depends on the AI model and which processor the AI inference is performed on, but in general, yes.

For low latency AI inference, we recommend to performing it on the NPU, which can handle int8 or int16 models. Most likely, available models today need to be quantized for Qualcomm's NPU, which can be done using SNPE or Qualcomm AI Hub.

As long as the model framework is one of following, the quantization and calibration can be performed to use them on the NPU.

  • TensorFlow
  • TensorFlow Lite
  • LiteRT
  • ONNX
  • Pytorch

Note that not all operations are supported. Please contact us for more details.

What AI modes can the EP-200Q support?

What operating system can the EP-200Q support?

Yocto Linux is the default support package from Silex.

We provides SDKs based on Yocto Linux. There are two SDKs from Qualcomm:

- Intelligent Multimedia SDK (QIMP)

  • This is the default SDK package that Silex will provide for EP-200Q customers.
  • The SDK is suitable for embedded devices such as AI cameras, tablets, mobile handheld devices, PoS systems, access control panels and other devices that do not require  real-time control of other systems.

- Intelligent Robotics SDK (QIRP)

  • This SDK is suitable for robotics or vision AI systems that interact with other controllers in real-time.
  • The SDK supports ROS2.

Silex supports customers in enabling Android devices on a case-by-case basis. Please contact us for further information about Android support.

How can I get started with the EP-200Q?

Silex provides the EP-200Q evaluation kit, EP-200Q-EVK.

It will come with pre-built firmware. We also provide the SDK for customers prototyping using the EP-200Q-EVK.

For software development, an Ubuntu 22.04 computer is required to compile firmware or application programs.

For more details, download the EVK introduction from the link below.

DOWNLOAD

Why Silex

The best customer experience for "When It Absolutely Must Connect".
partnerships

Best-in-class partnerships

Silex partners with best-in-class technology providers to offer proven, comprehensive solutions.
Faster

Wireless expertise

20 years of Wi-Fi and Bluetooth experience and exclusive access to Qualcomm technology.
support

Dedicated support

Local engineering resources for pre-sales consultation, integration, and post-sales support.
Unrivaled

Unrivaled quality

Highest reliability for medical and industrial applications.

Need help to get started and begin evaluation?

Learn more