-
Join TU Delft and work together with NXP to build low-power AI accelerators for self-healing analog/RF calibration, fixing noise/offset. Co-design algorithms & hardware and validate on real silicon
-
responsibilities will include: Design and develop analog circuits for implementing ONNs for computing. Modeling, simulate and benchmark different computing tasks such as sensor data processing. Explore ONN
-
related discipline. Strong background in analog and mixed-signal integrated circuit design. Experience with Cadence tools for schematic design, layout, and circuit simulation. Interest or prior experience
-
analog circuits for implementing ONNs for computing. Modeling, simulate and benchmark different computing tasks such as sensor data processing. Explore ONN implementation topology and its energy efficiency
-
for implementing ONNs. Modeling, simulate and benchmark different computing tasks such as combinatorial optimisation tasks and solving partial/ordinary differential equations with ONNs. Design and tapeout ONN chips
-
without neurons in physical systems, Ann Rev Cond Matt Phys14, 417 (2023) [4] Dillavou, Beyer, Stern, Liu, Miskin and Durian, Machine learning without a processor: Emergent learning in a nonlinear analog