In 2059, the computer under your desk will not exist. In fact, the very definition of a computer will change, as will the gadgets we carry around, the data centres used for storing information and the electronics in our living rooms.
Stan Williams, the noted HP researcher, says that technology changes by a factor of 100 every 10 years. In the movie 2001: A Space Odyssey, the HAL computer used something called "optical storage" – a medium so commonplace now they are used for throwaway marketing.
"In a few decades from now, our current concepts of a PC and IT in general will seem so quaint and arcane that the only place to see them at all will be in old attics or museums, where they will sit next to an abacus and vacuum tube computers, and seem very alien to those being born now," says Williams.
In fact, most researchers believe the technology we will use in 2059 has not been invented yet. Imagine the scenario: those who were born in the last 10-15 years have grown up with the Internet and computers.
They are what Gartner Group calls the "natives" (and the rest of us are immigrants). What will they invent when they are 50 and become the leaders of their generation? But despite this, even today there are some glimmers that showcase how technology might look in 50 years.
1. Interconnects will use photonics
According to Stan Williams, photonics will change the very definition of a computer. "Fifty years from now, any time information of any type is transmitted over a distance of more than a millimetre, it will be carried by photons," says Williams. The technology uses light to move data, and is already used in fibre optics to connect two server racks. In the future, photonics will be used for PC interconnects and will even be used inside the CPU. That means there will be no need for copper or metal inside a PC. In the far future, there's no reason why a computer could not exist in "free space" with no confines.
2. Zettascale computing will solve complex problems
In 10 to 20 years, supercomputer design will finally move to exascale, or millions of trillions of operations per second. According to Mark Seager, Chief Architect of HPC development at the Lawrence Livermore Lab in California, even the move to exascale (one thousand times faster than a petaflop) will require inventing new technologies, especially related to storage, power, and interconnects. A move beyond that - to zettascale (1,000 times faster than exascale) - will also require paradigm shifts.
Yet, the benefits are clear: protein weight measurements for cancer research will be easier, global climate analysis will help researchers predicate warming trends more accurately, and the most complex simulations - such as those that determine the effects of a global nuclear calamity - will be possible.
3. Computers will mimic the brain
IBM has already started a research project in its Almaden research centre to create a computer based on the human brain. As you can imagine, the processing power to mimic the brain will be immense - likely using supercomputers that do not exist yet. The idea is to use fibre channels in the same way as the brain - the microscopic strings that connect one part of the brain to another, and can interoperate at the same time.
This will require new programming paradigms, making use of successive and concurrent channels all at the same time, using multi-core CPUs and redundant storage arrays. Other institutions - including the Allen Institute in Seattle - have also started mapping the brain using high-res mapping techniques.
4. Electronics will charge wirelessly
Already, companies such as Convenient Power and Wi-Power are working on prototypes that can charge your smartphone with no wires. Yet, in 50 years this idea will be commonplace. In 2059, it's possible that all gadgets, computers, and electronic devices - and possibly even your home and car - will be charged by wireless power.
Wireless power works by transmitting an electric pulse as a frequency that only a specific receiver can read. In the US, NASA once proposed wireless charging satellites that would span the globe, but the idea never went anywhere. Yet, wireless charging makes sense because it reduces cables - in the same way that Wi-Fi replaces Ethernet.
Your comments (6) Click to add a new comment
budryerson
March 30th 2009
6. Dragon 10 "works better than it did a few years ago"? I should hope so. "Eventually" the PC will understand anything I say? "Eventually" is a lonnng time from now.
"Being a geek won't matter" because, if a technology is designed correctly, no one needs to be a "geek". They'll just use it.
Alert a moderator
dangrabham
March 30th 2009
5. Hi anon654, this has been updated. It should have read millions of trillions per second.
Thanks,
Dan
Deputy Editor
TechRadar.com
Alert a moderator
matrixdweller
March 30th 2009
4. In 50 years computing will be vastly different. There are a lot of technologies that haven't been invented yet that will shape it. Once you start predicting the future of computing past 20 years the probability of error goes up exponentially.
As for everyone will be geeks. Maybe compared to today's standards, but you'll always have your UberGeeks. Not everyone will be a brilliant programmer or get their kicks overclocking their liquid helium cooled massively multi-cored photonic CPUs in 2059.
Making the assumption that everyone will be geeks would be like saying, back in the early 20th century, that everyone would be an auto mechanic since automobiles would proliferate to a great extent. The more concise conclusion would be that everyone would know how to drive.
Sure the technical know how of the average person will far exceed that of the average person today. Similarly on how each generation's technical know how is more advanced than the preceding generation's (my parent's still can't set their digital clocks). There will always be those that have a more advanced comprehension of a subject than the general population.
Alert a moderator
pimlicosound
March 30th 2009
3. Being a geek won't matter? Surely there will still be people who know and care more about the technology they use than others. Sure, in 50 years social networking might not be geeky, but something else will come along - all new and technical - and that will probably be considered geeky.
Also, this whole article is assuming that we somehow get out of the downward spiral in education and engineering aptitude and end up producing some good inventors.
Alert a moderator
pete_l
March 29th 2009
2. This guy hasn't learned the lessons of history.
Let's look at lightbulbs: invented 130 years ago - still basically the same in application and format as they were when first introduced into houses 100+ years ago.
Or how about the telephone? Just as old as the bulb and still effectively the same as it's predecessors.
Or the car? 4 wheels, seats, same-ish size, faster and more reliable but still recognisable for someone from the 19th century.
The common element is that things PEOPLE interact with don't change much after their inception, because the people don't change much. Electrical appliances have to be of a certain size for us to use. So it will be in the future. Things with screens will still be recognisable as TVs or whatever, keyboards will still be keyboards (because the alternatives - which have also been around for 50 years - are so bad).
We do invent new things, and they can supplant old things, but proven design is permanent.
Alert a moderator
anon654
March 28th 2009
1. 'In 10 to 20 years, supercomputer design will finally move to exascale, or trillions of operations per second'
Isn't trillions of operations per second already possible? ie TERAflops. I think it should have said quintillions of operations per second because that is what an exaflop would be.
Alert a moderator
Tell us what you think
You need to Log in or register to post comments