Building analog neural circuits

This isn’t directly D-Wave related, but what these guys are working on is sufficiently similar in some important ways to our project that I thought I’d link to it. It’s also pretty cool.

Here is a link to an MIT tech review article:

Building a brain on a silicon chip

brain_x220

The basic idea in both this & our approaches is to create a grid of interconnected devices, where the interconnects set a pairwise coupling strength and the devices encode the information ultimately read out. There are very similar I/O, programming and integration issues between the two approaches. The big advantage most CMOS-based approaches have over us is that they get to use a $10B TSMC fab, while we had to develop our own foundry. It would be very interesting to see if our web services programming approach could work with analog neural net architectures…

5 thoughts on “Building analog neural circuits

  1. I remember this been tried about 20 years ago, with the Transputer(?) technology.
    Did not work, apparently.
    Better luck this time..

    • The only problem with that line of reasoning, is that it did work.
      And still does I believe, unless of course some rather fundamental
      properties of mathematics have changed in the past 10 years.

      In fact it may even work better now, you know… faster, more
      accurate. Anyway, who? can argue with nature, math,.. that sort
      of thing.

      By the way, really enjoyed your 3rd video presentation “Does an Explanation
      of Higher Brain Function require reference to Quantum Mechanics.
      The numerical analogy fits, if you insist on forcing it… and why not ?
      Has certainly created a big pot of boondogel money for academia, and you folks.
      However, more than a little has been lost trying to force a square peg into a
      round hole.

      Just wanted to ask, Were you guys on Ayahuasca when you thought all this stuff up ?

  2. A little off topic, but Luke watched some videos on quantum physics
    and deduced that since when observed, electron particles behave differently, they must be sentient. “They have awareness”, he said. Weird. Space: the illusory backdrop for quantum particles that are in fact always ‘sensing each other’, and when ‘watched’, they react. I can’t help but wonder what this might mean in the context of prions and viruses. All particles in superposition to each other, at all times. Awesome.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s