Salma Afifi

About

Salma Afifi is a pioneering researcher at the intersection of photonic computing and graph neural networks (GNNs), a field critical for modeling complex relational data in applications from drug discovery to social network analysis. Her most notable contribution is the GHOST architecture—a Graph Neural Network Accelerator using Silicon Photonics—which addresses the fundamental bottleneck of accelerating GNN inference. By leveraging the speed and energy efficiency of photonic circuits, Afifi’s work proposes a radical departure from traditional electronic accelerators, enabling faster and more scalable processing of graph-structured data. Though early in its citation impact (with 12 and 2 citations for the 2023 paper), GHOST represents a forward-looking synthesis of hardware design and machine learning, positioning Afifi as a key voice in the emerging field of neuromorphic photonics. Her research holds promise for real-time recommendation systems, robotic perception, and large-scale network analysis. As a researcher bridging optics and AI, Afifi is shaping the next generation of specialized hardware for the most computationally demanding deep learning workloads.

Research Focus

Key Achievements

Top Papers

1. 1
   GHOST: A Graph Neural Network Accelerator using Silicon Photonics

   12 citations · 2023
2. 2
   GHOST: A Graph Neural Network Accelerator using Silicon Photonics

   2 citations · 2023

Key Collaborators

Contact & Links