Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?

Submission + - Computing in the new world: Scaling to 1e6+ Cores ( 2

mattaw writes: In my* blog post I describe a system designed to test a route to the potential future of computing. What do we do when we have computers with 1 million cores? What about a billion? How about 100 billion?

None of our current programming models or computer architecture models apply to machines of this complexity (and with their corresponding component failure rate and other scaling issues). The current model of coherent memory/identical time/everything can route to everywhere just can't scale to machines of this size. So where did the scientists** at the University of Manchester (including Steve Furber one of the ARM founders) and the University of Southampton turn for a new model? They took one straight out of their own heads. Quite literally: their brains.

Our brains just don't work like any computers we currently make. Our brains have a lot more than 1 million processing elements (more like the 100 billion), all of which don't have any precise idea of time (vague ordering of events maybe) nor a shared memory and not everything routes to everything else but anyone who argues the brain isn't a pretty spiffy processing system ends up looking pretty silly (assuming they have one). In effect modern computing bears as much relation to biological computing as the ordered world of sudoku does to the statistical chaos of quantum mechanics.

Read the article*** to see a preview of the brain turned into hardware (and of course you will read all the papers from Manchester's website before posting won't you). Who says science is boring?

* Note the subtle declaration of interest. I also work at the University of Southampton, UK.
** I am not and have never been one of these mighty people trying to change the world. I claim no credit. At best I helped some of the PhD students and staff with a few bits and bobs plus the odd ARM development kit.
*** No free lunch here /. You may have to actually read the source article.

This discussion was created for logged-in users only, but now has been archived. No new comments can be posted.

Computing in the new world: Scaling to 1e6+ Cores

Comments Filter:
  • Just how much SMP traffic is passed around a system at 64 cores? I *think* it's something 1.3Tb/second. What frequency do we expect the core interconnects to be running at to sync the required SMP data?

Every successful person has had failures but repeated failure is no guarantee of eventual success.