2009: UCSD ISNL Neural Broadcast

DARPA announced the BAA for the Synapse Program.  The idea was to build the electronic equivalent of a mouse brain and then a cat brain using electronics and nanotechnologies.  This was just one of many projects that brought a brainstorming team to the lab often where I worked (ISN).  As always my concern was the connectivity and how to deliver neural event messages to their intended destinations on time without excess jitter.  This was a major system integration issue, about as important as figuring out how to do synaptic updates in place.  It turns out that understanding the ‘connectome’ of the brain is absolutely key to understanding how it works at all (e.g. see Swanson).  The reason for this is that the brain is all about encoding relationships in the form of what can be inferred from what.  Relationships (connections) are at the foundations of patterns recognized and produced.  These in turn enable the structure in thought, language, movement and all manner of adaptive anticipatory behavior.

Just before leaving the lab, due to lack of funding again, I helped get a project going to simulate a system for broadcasting events.  People did not seem to have much confidence in my ability to write code.  So I decided to prove them demonstrably wrong.  I shut my door and over a two week period I taught myself Python and SimPy.   I simulated message traffic in an architecture we were entertaining as part of a neuromorphic design.  The day that I left, I ran it for them.  I still remember the incredulity displayed.  🙂   Subsequent refinements by others eventually resulted in another publication on Scalable Event Routing.