Aardvark Daily

Please visit the sponsor!

From sand to the stars

It can be very easy to forget just how impressive modern technology has become.

Thanks to the power of the microcomputer, many otherwise incredibly complex devices and the tasks they perform have become very simple.

Want to wash the clothes? Biff them in the washing machine and hit the "Auto" button. If yours is one of the newer "smarter" machines, it'll work out how much water to use, how long to wash, rinse, spin and how may times to repeat the process in order to achieve a totally satisfactory result.

Want to record a programme on the VCR? Just whack a G-code in and the machine will work out for itself what time, what channel, and how long to record for.

Feel like taking a picture? Just turn on the digital camera, point it at the subject and push the button. Exposure, shutter speed and a myriad of other variables will be calculated and adjusted in the blink of a shutter.

And just what makes all this fancy stuff so simple?

Well its a device that's just had its 60th birthday.

I'm talking about the good old transistor of course.

Thanks to the boffins at Bell Labs, we were able to ditch the world of glowing filaments, humming HV transformers, and bulky, hot racks of valves -- in favour of tiny solid state semiconductor junctions made from sand.

And, in the evolution of mankind, the 60 years it's taken to turn the transistor from a novel electronic component into the cornerstone of almost all modern technology is incredibly short.

Moore's Law has really come into play and the world is now filled to overflowing with large-scale integrated circuits, each with thousands or millions of tiny transistors inside.

Please visit the sponsor!

Did you realise that today's PCs have more bits of solid-state memory in them than there are people on the face of the planet?

Yep, a PC with 1GB of RAM has over 8 billion bits of storage capacity, an absolutely stunning amount.

Unfortunately, we seem to be reaching the practical limits for transistor technology and now we're finding that it may be hard for Moore's Law to remain valid.

Processors are now running so fast that even with our most dense packing of transistors on the silicon wafer, the time it takes for electricity to flow around the conductors on that wafer is becoming a limiting factor.

Believe it or not, electrical signals (and light) only travel 100mm during the single cycle of a 3GHz computer's clock. Given the speed of light, it becomes apparent just how fast a GHz really is.

Fortunately, the limitations of conventional transistor-based processing and memory systems may soon be overcome (about 10 years they reckon -- ha ha) thanks to quantum computing.

If/once we are able to harness the laws of quantum mechanics to side-step the restrictions imposed by conventional physics, we open up a whole new world of processing power that will make today's supercomputers look like that humble first transistor of 1947.

What's more, given the rate at which technology tends to progress on a decidedly non-linear scale, chances are that, within a decade of the first practical quantum computer being developed, we'll be several orders of magnitude further down the road.

But just what will we use all this unimaginable computing power for?

Imagine being able to model the entire planet's weather systems for weeks ahead with very high degrees of certainty -- and perform the necessary modelling in less than a second -- woo hoo!

How about being able to simulate/model almost any device or idea you can conceive, right down to a molecular level -- and again, do it all in the blink of an eye.

Might this kind of computing power open the door to truly practical interplanetary and pan-galactic space travel by exploiting other aspects of quantum mechanics?

Could the sheer computing power that quantum computers might offer us finally result in some true artificial intelligence?

What will happen if we build a quantum computer that models the brain's neural/synaptic design, not only in form but also in sheer scale (something totally impractical with existing technology)? Will we end up creating a new thinking, reasoning, intelligent entity? How would we handle the ethics associated with such an achievement?

The truth is that we'll almost certainly find limitless applications for effectively "unlimited" computing power -- just as we've found a mountain of applications for the humble transistor and its great grandchildren.

So today's questions are:

Apart from the obvious (ultra-secure encryption) what would we actually use a quantum supercomputer for?

If we have the ability to create a true sentient artificially intelligent machine with consciousness -- should we? And if we did, would turning it off be an act of murder?

Or will truly practical quantum computing remain as elusive as fission reactors and cheap fuel cells?

Have your say on this...

PERMALINK to this column

Oh, and don't forget today's sci/tech news headlines

Rank This Aardvark Page

Change Font