One of the things that most annoyed me about the article on graphics was the author’s subsequent attempts (once he came under criticism) to justify bad technological advice through the excuse that falling hardware prices meant that his advice to use inappropriately optimised formats did not matter.
For the truth is that cheaper hardware expands the use of technology at both the “fat” end (ie your 4GB quad core desktop) and the “thin” end – your mobile phone and lower still. As computing becomes ubiquitous then appropriate optimisation will be as important as it ever was.
And appropriate optimisation means efficient and effective algorithms.
In fact, suggests a report quoted here, improvement in algorithm efficiency has contributed more – by an order of magnitude – to advances in some branches of computing than falling hardware prices (I am using price here as an analogue for the general Moore’s Law advance).
The report only appears to cite two specific examples of how algorithmic improvement has far outstripped the silicon, so I would be wary of claiming in general that algorithms have contributed a 30 – 43 times greater improvement to computing efficiency in the 15 years between 1988 and 2003 (something not bothering the slashdot summariser, so this is now likely to become the myth), it is a remarkable finding nonetheless.
cartesian product ... stuff about computing mostly 