Position @ The University of Edinburgh, United Kingdom
PhD Student, In-Memory Computing Architectures for AI
The project aims at exploring the potential promise of the In-Memory Computing (IMC) concept to target the bottlenecks of AI hardware. Digital IMC is proposed to bridge the Von-Neumann performance gap for AI applications where massive data workloads are consumed. However, the conventional binary computing domain degrades the benefits of IMC due to its computational complexity. The project targets exploring different unconventional computing domains (like Stochastic and Quasi-Stochastic) for IMC. Emerging technologies, with a specific focus on RRAMs, are proposed to increase the on-chip computing memory capacity.
The required skills are as follows:
* Mixed-Signal IC design using Cadence Tools (mandatory).
* Previous experience in Tape-outs and Chip Testing.
* RRAMs and/or other Emerging Devices.
* TCL and Makefile scripting.
For informal enquiries, please contact Dr. Shady Agwa, email: shady.agwa@ed.ac.uk
#J-18808-Ljbffr