I had lunch two weeks with two old friends and when I mentioned that we were working with a new company, TTM, helping them to exhibit and present at Semi-Therm last week and one of them told me an interesting story.
He writes microcode and debugs complex system design issues for a storage area network company. He is comfortable with both hardware and software and for some reason he didn’t elaborate on he had swapped a CPU from his son’s PC with the one in his. He detached the heat sink to be able to remove it from his son’s computer. Just as he touched the power on button on his PC he realized that he had forgotten to re-attach if after installing it in his computer. Alas, before he could even take his finger off the button smoke was coming out of the computer: it died instantly.
Needless to say designing the thermal network that protects a chip, a system, an LED lighting system (which these days can be an overhead light, a streetlamp, or even a television screen), or an electric vehicle is more important than ever.
I think there is another lesson for entrepreneurs in this, so often we look at the primary driver for system performance:
- transistor size
- microprocessor clock speed
- miles per gallon
- zero to 60 in N seconds
that we overlook when secondary constraints start to become as or more important:
- battery life for portable devices
- memory bus width or instruction word length (E.g. VLIW) so that the same cycle does more work
- switching to hybrid power trains
- 60 to zero: highway crash survivability
Christensen addressed how performance overshoot enables new markets in “The Innovator’s Dilemma” I think there is another category of opportunity that’s “the hill behind the hill” or addressing the problem that an industry has promoted when they have solved a primary problem.
Pingback: SKMurphy, Inc. Moore's Law Enables New Uses For Old Algorithms - SKMurphy, Inc.