Many years ago, I heard it said that, at the time of the introduction of the steam engine in England, the changes over just one generation were so profound that a parent born before the Industrial Revolution had more in common with someone from the Stone Age than they did with their own children. With technology moving so fast, it’s sometimes hard not think that the same thing is happening now. Maybe the changes aren’t quite as profound – as someone who was born during the era when computers became a daily reality, I still do have plenty in common with my digital native kids. (They may think of me as Stone Age, but I know better.)
Major changes do seem to be occurring more rapidly than they did when the steam engine led to the emergence of a manufacturing economy.
These musings came to mind when I saw an article on neuro computing on EE Times.
The article was about IBM’s “brain-like chips”, which are being used at Lawrence Livermore National Laboratory (LLNL) “for simulating the deterioration of our aging nuclear arsenal—currently the most difficult problem for supercomputers to solve worldwide.”
The new neuro computing technology will be concentrated in areas of importance to the National Nuclear Security Administration (NNSA): cyber security, nuclear deterrent, non-proliferation.
According to Dharmendra Modha, IBM fellow and chief scientist for brain-inspired computing at IBM Research-Almaden:
“NNSA’s Advanced Simulation and Computing (ASC) program will evaluate machine learning applications, deep learning algorithms and architectures plus conduct general computing feasibility studies. ASC is a cornerstone of NNSA’s Stockpile Stewardship Program to ensure the safety, security and reliability of the nation’s nuclear deterrent without underground testing.” (Source: EE Times)
IBM’s neuromorphic computer packs an awful lot of punch in a small form factor, and with low power consumption:
“The 16-chip neuromorphic system represents 50 times the computing power of today’s computers,” Modha told us. And consumes just “1/10th the power of a dim lightbulb.”
This neuro computing – which IBM has branded True North – deploys “deep learning supported by hardware that emulates 16 million neurons and 5 billion synapses.” IBM believes that their approach will prove better than massive multicores and other processing schemes.
The article gets into some detail on how TrueNorth works, and is definitely worth a read. (There’s also an info-graphic if you don’t have time to read the full piece.)
I’ve seen an awful lot of technology changes in my lifetime. With our smartphones, we’re all carrying the equivalent of yesterday’s mainframe in our pocket. And there’s no underestimating the changes that have been brought about by the growth of social media. At Critical Link, we’re constantly seeing changes in computing power, size, and power consumption, and incorporating these changes in our SoMs.
But “50 times the computing power” and “1/10th the power of a dim lightbulb”?
Easy to see how you can start to feel like you’re a Stone Ager!