Irregular Webcomic! #733

I never had any intention of implying that the Minoan labyrinth on Santorini and the "maze of twisty little passages, all alike" encontered by the space crew in cyberspace were in any way related, but so many readers spotted that I reused the same maze set for both of them that I thought, "Why the heck not? When else am I going to get an opportunity like this for a crossover between these themes?"

2014-03-03 Rerun commentary: For those of you paying attention, Serron's mention of "microseconds" translates to roughly the same thing as Monty's "miles". Light, or a signal in an electrical circuit (like the cyberspace realm in which they are currently embedded), travels close to 300 metres in a microsecond, which is roughly of the same order of magnitude as a mile. A mile is 1609.344 metres.

That's an exact conversion, by the way, not just an approximation to three decimal places. The mile has been defined in terms of SI (metric) units since 1959.

EDIT: Forum poster x-viila says:

Signals don't actually travel at c - because that is of course the speed of light in a vacuum. Cables aren't vacuums, even optical cables have high refractive index optical medium. Good approximation is that signals move at roughly two thirds of c plus/minus, copper maybe a little higher, optic maybe a little lower, but varies depending on the exact types of cables.
It's quite amazing that modern computers work at all with their multi-gigahertz clock frequencies. Even if we're generous and call it c, at 3 GHz the signal can physically move just under 10 cm in a clock cycle. And that is just getting from A to B, not counting all the inductive and capacitive effects of the lines and transistors between!
Any wonder that the original Pentium 4 that ran at these kinds of frequencies had a ridiculously long 40+ stage pipeline? :P That gives you 40 cycles to push the signal through all the work the CPU needs to do on it. But the pipeline also lets 40 instructions percolate through it at the same time so each effectively executes in "single cycle". The problem comes when the line stalls due to data-hazards (instruction needs result that hasn't been computed by earlier instruction yet) or branch misses (CPU followed wrong branch (because real result wasn't calculated yet) and has been speculatively executing wrong instructions in the pipeline and has to flush it)... With 40 instructions working through the pipeline at any one time both of those were pretty common cases.
This also demonstrates why we've stopped pushing the gigahertzes ever higher and instead work on making the CPUs wider by adding cores even though that's much harder to program for...