Sources Say Apple Originally Planned AMD Chip For MacBook Air 197
Several media sources (here's PC Magazine's version), all seemingly based on an account at SemiAccurate citing (but not naming) "multiple sources," report that Apple originally planned an AMD-chip based MacBook Air, rather than the Intel-based version that emerged later ("Plan B," says the report).
AMD always considered ... (Score:5, Interesting)
Re:This would have been great for.. (Score:2, Interesting)
I was running os x on my old amd system quite early on in the osx86 scene...
Re:Intel (Score:5, Interesting)
At least AMD doesn't build hardware level backdoors into their CPUs.
That you know of.
Re:Intel (Score:4, Interesting)
Re:In summary (Score:5, Interesting)
Re:CPU & GPU performance not relevant (Score:5, Interesting)
People tend to conflate power with energy, and you may be doing it here. If you're going to be executing a particular job, and you want to optmize its efficiency, then it will consume some power over some time period, which is ENERGY. On the other hand, if you're talking about the battery life of your laptop, then the computer is almost completely idle, and what we want to therefore minimize is idle and average power.
Optimizing just for power isn't sufficient. If something uses half the power but takes 4 times as long, then it's twice as bad. However, we don't typically wake our computers to run compute-intensive jobs, just to put them back to sleep when those are done. We do a lot of screen-staring, which complicates the issue.
Interestingly, performance per watt IS in the right units. Performance would be something comparable to operations per second, while watts is joules per second. The seconds cancel out, giving you operations per joule, which is the correct efficiency metric.