There's nay sayers who think 50Ghz machines are impossible, there's people who say processors cannot physically get any faster ... They've been saying that since the 100Mhz chips.
What they're forgetting is you can already buy a microprocessor based computer with over 8000 dual core CPU's installed. Sure you cannot fit it into your bedroom right now but if there's one thing you can rely on in technology it's that things get smaller.
There's also not much hope of being able to afford one for the time being because i'm talking about a machine manufactured by Cray - all the same the point is technology exists and there will always be a financial reward for the company who can build a faster computer.
If manufacturers really cannot get any more speed out of machines they'll simply double them up, as evidenced by the fact we've long since been able to buy multi-processor machines and now we can get multi-processor processors...
Cooling is the big constraint at the moment, and yet the mainstream solutions still dont have dedicated cooling units - just a traditional heatsink and fan not much more advanced than we had in the 486 days. Overclockers have been doubling/tripling processor performance for years and if their methods go mainstream the solutions will become both more elegant and cheaper.
It's more than likely we'll start thinking of computer performance in a different measurement than Ghz before long, as more expanded commandsets bring speed boosts for specific tasks by delegated circuitry we're likely to see processor speeds measured by benchmarks, subdivided into tasks.
Networking speeds will see the biggest boost in the next 10 years, the internet will reach the state where it is faster to save your document to a server in Taiwan than it is to write to your floppy disk ... oh wait we've already gone past that point ... :)
As the internet gets faster and more secure* (Microsoft dependancy here, the internets reputation and consumer confidence depends upon the outcome of Vista's malware defences more than ever before as continued security breaches are currently somewhere beyond the description of "epidemic") more non-computing technologies will be internet reliant.
We'll likely loose broadcast television within the next decade or two - supply on demand is the future there and I expect improved networking will make it far more powerful than the current "Sky+" we get offered in the UK. The commercial sucess of true supply on demand will depend more than anything on who buys the rights, the only people who can afford Sky+ at the moment are Managing Directors and council flat tennants.
What they're forgetting is you can already buy a microprocessor based computer with over 8000 dual core CPU's installed. Sure you cannot fit it into your bedroom right now but if there's one thing you can rely on in technology it's that things get smaller.
There's also not much hope of being able to afford one for the time being because i'm talking about a machine manufactured by Cray - all the same the point is technology exists and there will always be a financial reward for the company who can build a faster computer.
If manufacturers really cannot get any more speed out of machines they'll simply double them up, as evidenced by the fact we've long since been able to buy multi-processor machines and now we can get multi-processor processors...
Cooling is the big constraint at the moment, and yet the mainstream solutions still dont have dedicated cooling units - just a traditional heatsink and fan not much more advanced than we had in the 486 days. Overclockers have been doubling/tripling processor performance for years and if their methods go mainstream the solutions will become both more elegant and cheaper.
It's more than likely we'll start thinking of computer performance in a different measurement than Ghz before long, as more expanded commandsets bring speed boosts for specific tasks by delegated circuitry we're likely to see processor speeds measured by benchmarks, subdivided into tasks.
Networking speeds will see the biggest boost in the next 10 years, the internet will reach the state where it is faster to save your document to a server in Taiwan than it is to write to your floppy disk ... oh wait we've already gone past that point ... :)
As the internet gets faster and more secure* (Microsoft dependancy here, the internets reputation and consumer confidence depends upon the outcome of Vista's malware defences more than ever before as continued security breaches are currently somewhere beyond the description of "epidemic") more non-computing technologies will be internet reliant.
We'll likely loose broadcast television within the next decade or two - supply on demand is the future there and I expect improved networking will make it far more powerful than the current "Sky+" we get offered in the UK. The commercial sucess of true supply on demand will depend more than anything on who buys the rights, the only people who can afford Sky+ at the moment are Managing Directors and council flat tennants.