It remains one of the biggest problems for many who use computers. It is supposed to be getting better, but statistics from full time usage shows it just is not so.

What is the problem? Hardware failure. In his post this morning, Robin Harris shows that hardware failures are no less than 10 years ago. Does this speak ill of the quality control of today, or is it that we build (somehow) to a standard of acceptable failure? No substantive proof is given for either assertion, but it is rather disconcerting that in the 21st century we must still deal with things that cause such major problems, without much hope of improvement.

Of course, mechanical things will always wear out – no one has been able to remove friction and heat build up from moving parts – but things like CPUs, memory, and graphics chips should have become much more reliable by now – especially since the number of ‘experimenters’ has decreased in percentage, since every Tom, Dick, and Mary now owns a computer.

Good news for software companies though – it is still possible to blame hardware and be right, for the most part.

-

Technorati Tags: computer failures - CPU - memory - graphics chips - reliability - large scale testing - acceptable failure