Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Neuromorphic Computing - An Overview

Benedikt Jung, Maximilian Kalcher, Merlin Marinova, Piper Powell, Esma Sakalli

TL;DR

The paper surveys neuromorphic computing as a brain-inspired alternative to von Neumann architectures facing memory–processor bottlenecks and Moore’s law limits. It compares three technological avenues—spiking computations on conventional GPUs, dedicated neuromorphic chips (e.g., Loihi) and photonic systems (LNPS, VCSEL-based)—highlighting their respective energy efficiency, speed, and scalability tradeoffs. Key insights include the event-driven, analog/mixed-signal nature of neuromorphic chips and the ultra-fast potential of photonic platforms, with reservoir computing enabling efficient training. The work emphasizes that while neither platform is yet mainstream, neuromorphic hardware could become a foundational technology for future energy-efficient AI and real-time sensing applications.

Abstract

With traditional computing technologies reaching their limit, a new field has emerged seeking to follow the example of the human brain into a new era: neuromorphic computing. This paper provides an introduction to neuromorphic computing, why this and other new computing systems are needed, and what technologies currently exist in the neuromorphic field. It begins with a general introduction into the history of traditional computing and its present problems, and then proceeds to a general overview of neuromorphic systems. It subsequently discusses the main technologies currently in development. For completeness, the paper first discusses neuromorphic-style computing on traditional hardware, and then discusses the two top branches of specialized hardware in this field; neuromorphic chips and photonic systems. Both branches are explained as well as their relative benefits and drawbacks. The paper concludes with a summary and an outlook on the future.

Neuromorphic Computing - An Overview

TL;DR

The paper surveys neuromorphic computing as a brain-inspired alternative to von Neumann architectures facing memory–processor bottlenecks and Moore’s law limits. It compares three technological avenues—spiking computations on conventional GPUs, dedicated neuromorphic chips (e.g., Loihi) and photonic systems (LNPS, VCSEL-based)—highlighting their respective energy efficiency, speed, and scalability tradeoffs. Key insights include the event-driven, analog/mixed-signal nature of neuromorphic chips and the ultra-fast potential of photonic platforms, with reservoir computing enabling efficient training. The work emphasizes that while neither platform is yet mainstream, neuromorphic hardware could become a foundational technology for future energy-efficient AI and real-time sensing applications.

Abstract

With traditional computing technologies reaching their limit, a new field has emerged seeking to follow the example of the human brain into a new era: neuromorphic computing. This paper provides an introduction to neuromorphic computing, why this and other new computing systems are needed, and what technologies currently exist in the neuromorphic field. It begins with a general introduction into the history of traditional computing and its present problems, and then proceeds to a general overview of neuromorphic systems. It subsequently discusses the main technologies currently in development. For completeness, the paper first discusses neuromorphic-style computing on traditional hardware, and then discusses the two top branches of specialized hardware in this field; neuromorphic chips and photonic systems. Both branches are explained as well as their relative benefits and drawbacks. The paper concludes with a summary and an outlook on the future.

Paper Structure

This paper contains 9 sections, 2 figures.

Table of Contents

  1. Introduction
  2. Introduction to Spiking Systems
  3. Technologies
  4. Neuromorphic Computing on Conventional GPUs
  5. Neuromorphic Chips
  6. Photonic Systems
  7. Lasers
  8. Other Photonic Systems
  9. Conclusion

Figures (2)

  • Figure 1: The compute density of modern GPUs. Taken from svedin in compliance with the figure's CC BY-NC-ND license allowing attributed free use.
  • Figure 2: Theoretical speed versus efficiency of neuromorphic photonic systems as compared to other neuromorphic systems. Taken from shastri2017 in compliance with the figure's CC BY-NC-ND license allowing attributed free use.