Work on a neuromorphic chip carried out at the University of Edinburgh in the UK has revealed that the biological learning mechanism ‘spike-timing-dependent plasticity’ (STDP) reduces the variation in performance between neurons arising from both mismatch in fabrication and inhomogeneities in the electronic design.“Biological neural networks have homeostasis built in at various levels, of which STDP is one form,” said Dr Simeon Bamford who carried out the work. “As we continue to build neural hardware we can expect to benefit from these effects more and therefore worry less about process variations than we might otherwise have thought.”