FrederikSchack
Regular
Implications for SNN
It's definitely interesting, but it's a research project at Polytecnico di Millano and probably far away from commercialisation. Just look at how long it took Brainchip from final design of Akida 1000, to gain traction and being designed into other products. We also have to consider that this is a first prototype and many more steps of refinement may be needed.
It's produced on top of a CMOS that is controlling the neural net, probably a good way to stabilize the thin and fragile 2D layer, but it probably makes it harder to integrate with other technology. It's apparently compatible with the existing production technology, which makes it much easier to get to market, than when you have to invent new technology for production.
I suppose that it's still several years into the future, before this hits the market, but it's worth to keep an eye on and see if there's any indication to the performance of the device.