Edge Computing

(Information Science Expert) Lecture on IoT and Edge Computing

• It is challenging to handle all computations by communicating with a server.
• Therefore, the approach is to perform computations at the edge.
  • Striving to perform tasks such as Approximation Calculation within limited energy and computational resources.
• Requirements for edge computers:
  • 1 is a major premise.
  • Additionally, 2 and 3 pose conflicting challenges, known as a trade-off.
    • Designing a CPU for versatility consumes a lot of energy.
    • Creating an application-specific processor for energy efficiency sacrifices versatility.
  • How to address this:
    • Approach 1: CPU + Accelerator
      • ASIP: A combination of a general-purpose processor and an application-specific accelerator.
        • The accelerator handles specific heavy tasks.
        • A blend of ASIC and GPP advantages.
      • Non-application-specific accelerators
        • Aim for higher versatility and energy efficiency than ASIP.
        • Example of use: Approximate Calculation in Image Processing (approximate calculation based on reusing computations).
    • Approach 2: Domain Specific Architecture