UltraTrail - Ultralow-Power AI Accelerator for Edge Devices

Image
UltraTrail_System_Architecture

UltraTrail is a configurable hardware accelerator for real-time inference of temporal convolutional networks (TCNs) on edge devices. Designed for near-sensor signal processing on energy-constrained platforms it features an optimized dataflow for one-dimensional convolution with a total power consumption in the low microwatt range. The parameterizable architecture combined with a hardware-aware neural architecture search (NAS) allows an automatic generation of domain-specific accelerator instances. Accurate models for power, performance, and area enable a fast design-space exploration. A UMA backend is provided to support network model deployment using TVM.

Visibility
Publicly available!
Date of Availability
Contact

Paul Palomero Bernardo
Sand 13
72076 Tübingen
Deutschland

Contact Email
paul.palomero-bernardo [at] uni-tuebingen.de
Asset Reference Description

UltraTrail comes with a UMA backend for DNN deployment.

Target TRL at the end of phase 1
TRL4: Versuchsaufbau im Labor --> Plattform mit Parametrisierung
Target TRL at the end of phase 2
TRL5: Versuchsaufbau in vereinfachter Einsatzumgebung --> Optimierte Plattform
Component is required for safety critical systems
Nein