Data Analytics: Time-Series Data and Regression Analysis

GPUS ARE INEFFICIENT FOR TRAINING SMALLER AI MODELS USED IN MANY INDUSTRIAL, COMMUNICATIONS, AND DATA ANALYTICS APPLICATIONS

For smaller AI models, GPUs are typically underutilized, and therefore FPGAs can bring dramatic performance boosts.  Furthermore, these small AI networks can easily be deployed at scale for training at the Edge,  whereas centralized training cannot scale to address each and every endpoint. Training at the edge with DeepAI alleviates these scalability issues. 
 

Deep-AI provides higher-performance integrated training and inference solutions, which also save significant power and cost over GPU based solution.

  • Supported Neural Networks:
  • TCN
  • MLP
  • Others