Jun
24
If Fuel Efficiency Gained Like Processing Power
June 24, 2008 | 7 Comments
Professor Steve Furber, professor of computer engineering at the University of Manchester had an idea to support his lecture last Friday. The lecture is entitled the “Relentless March of the Microprocessor” for the inaugural Kilburn Lecture to mark the 60th anniversary of the “Baby” computer built at the university in 1948. He compared the rate at which microprocessor power has improved to fuel economy.
In the lecture Professor Furber noted that processors have improved by 50 billion times since “Baby” in 1948. Baby was 2200 pounds and occupied an entire room. Now a processor is a chip smaller than a fingernail, so 50 billion seems plausible in mass, and surely the case in processing power. More interesting is that the “fuel efficiency” of processors has kept pace, too. Furber offers that the equivalent today if UK national fuel use improvements had kept a pace similar to processors a single liter of fuel could power the UK for a year and the world’s oil reserves would last the lifetime of the solar system.
 |
The Relentless March of the Microchip: The Kilburn Lecture Prof. Steve Furber Presentation from Manchester University, Manchester, UK 2008-06-20 12:00:00.0 IT Channel |
But that is just the set up for what comes next – Professor Furber discusses the coming use of micro processing to mimic biological processes. Furber says, “Biological systems demonstrate many of the properties we aspire to incorporate into our engineering technology . . .” He offers that scientists still don’t understand many of the principles of operation of the complex human mind, yet they are becoming powerful enough to model significant components of brain function.
Furber heads the SpiNNaker project at the University of Manchester that aims to build a massively-parallel chip multiprocessor for modeling large systems of spiking neurons in real time. The goal is building a machine that draws on a million ARM processors linked by a communications system that can achieve the very high levels of connectivity seen in biological neural systems. The plan envisions a computer processing capability of modeling a billion neurons in real time – about one percent of the human brain’s power.
But the illustrative point of the fuel use compared to processor improvements is the most interesting observation. While a 50 billion fold improvement isn’t practical on the way to 100% efficiency it does show that fuels and energy have been too cheap to focus our minds on the ways to get work done with lower fuel and energy inputs. Cheap has allowed us the luxury of ignoring the efficiency gains we will need at the consumption end. By no means are we any where close to getting to optimal efficiency form any source of fuel or energy, but $135 oil is a strong motivator to focus the mind – which is as Furber notes – is far more powerful than any computer.
It is a sweet and sour message from the good Professor, and when read or thought through the message is clear, there are solutions to expensive oil, we just have to apply our minds to getting it done – Faster, Better & Cheaper!
Comments
7 Comments so far
Feedburner
Reddit
Stumbleupon
I have enjoyed reading – thanks for sharing your story. Greetings.
very good information you write it very clean. I’m very lucky to get this info from you.
Excellent reading. I appreciate your effort, have a good day!
Interesting read, perhaps the best article iv’e browse today. We learn everyday cheers to you!
Hello, this is my first time i visit here. I found so many interesting in your blog especially on how to determine the topic. keep up the good work.
I was just having a conversation over this I am glad I came across this it cleared some of the questions I had.
I’ve been checking your blog for a while now, seems like everyday I learn something new 🙂 Thanks