RESEARCH
Scientific Focus
AATLAS develops hybrid analog–digital solvers to address NP-hard optimization problems such as Max-Cut, k-SAT, and the traveling salesman problem. The project explores how Field-Programmable Analog Arrays (FPAA) and Field-Programmable Gate Arrays (FPGA) can be combined to deliver scalable, energy-efficient computing architectures.
Work Package 1 –
Analog Solvers
Implementation of Hopfield/Ising-type networks on FPAA, with demonstrations on medium-sized problems.
Work Package 2 –
Hybrid Architecture
Integration of FPAA cores with FPGA modules for control, calibration, and pre/post-processing, enabling scalability up to ~1000 nodes.
Work Package 3 –
System Studies and Benchmarks
Comparative evaluation of the hybrid solver against digital methods, including complexity analysis and energy efficiency metrics.