Take advantage of a fully integrated hardware solution for general-purpose AI computing. The GPX series of software-defined AI processors enable inference and training on-device, in clients, and at-the-edge. Build intelligent products with on-board AI functions even when not connected to the cloud. Differentiate your AI applications with ultra-low power consumption, high speed, and flexible programmability. Deliver AI for everyone.

GPX-10 AI Processor

The GPX-10 software-defined AI processor is ideal for on-device inference and training workloads. The 10-core DigAn™ AI engine and highly integrated architecture of the GPX-10 provides high performance sensor fusion with ultra-low power consumption for inference and training tasks. The GPX-10 can be trained and customized in the field and can collect and label data so developers can improve their neural network models even when not connected to the cloud.

Key Features:
  • 10 AI cores provide 2560 MAC operations per clock cycle
  • ~80µW active power consumption in inference tasks
  • Sensor fusion with 8 I/Os
  • Dynamic Character Morphing between inference and training
  • Capture and Label data in real time
  • Years of operations on a coin cell battery
Download GPX-10 White paper
GPX-4 AI Processor

The GPX-4 is a streamlined AI inference processor in a smaller package. You can add ultra-low-power AI inference functions to any device in a small footprint and operate without connecting to the cloud.

Key Features:
  • 4 AI cores provide 1024 MAC operations per clock cycle
  • ~60 µW active power consumption in inference tasks
  • Sensor fusion with 8 I/Os
  • Years of operations on a coin cell battery
  • Lower cost than the GPX-10

Developer Tools for GPX Processors

Ambient SDK

Ambient Scientific provides tools for application developers to implement any neural network model and other non-AI functions onto GPX processors. The Ambient SDK lets you build your AI applications in popular languages then compile your code to your GPX processors. Build your own models and scale your IP across your GPX enabled devices with the Ambient SDK.

The Ambient SDK currently offers support for...

  • TensorFlow

… and will add support for

  • PyTorch
  • Caffe
  • Theanos
Ambient Training Vault

Harvest your data from the GPX-10 and program on-device training with the Ambient Training Vault. Under a subscription model, you can retrieve your device collected data and use it to personalize applications, improve application performance or scale your AI-enabled ecosystem. Differentiate yourself from your competitors and build new business models for your products with the Ambient Training Vault.

  • Label data as it is captured
  • Harvest your labelled data from the GPX-10 via the Training Vault APIs
  • Own your application’s data and use it in other AI applications