mult.ifario.us

Reading some coverage of the Republicon convention and seeing pictures of Bush on the big screen brought to mind the classic Apple 1984 ad, and the following visual analogy is obvious:

For what it's worth, I've never felt much of an affinity for either party, and while I voted American last election (i.e., against Bush), I can understand how people might identify with what Bush said. I just can't understand how anyone could support what he did.

In spite of the fact that my last Apple workstation failed rather ingloriously after only a couple of years of use, I went ahead and replaced it with another Apple workstation, an eight-core Mac Pro.

As an experiment, I decided to run the same Erlang benchmark (big.erl) that I ran on the quad-core machine, this time with Erlang R12B. The previous results showed that four schedulers was optimal on the four-core machine. Here are the results of the same test battery on the eight-core machine:

Two things are odd about this chart:

1. The running times appear to be about equal to the running times for the benchmark on the quad-core machine. The raw clock speeds aren't that different per core (2.5GHz G5 versus 2.8GHz Xeon), so maybe it's not unreasonable for that to be a draw.
2. Four schedulers appears to be the optimum (from the set {1,2,4,8,16,32}), where eight would have been the expected value.

It turns out that the optimality of four schedulers in this case doesn't disprove the hypothesis that the optimum number of schedulers equals the number of cores, since the benchmark only appears to be utilizing three of the eight cores:

The question is why the Erlang VM isn't using the available CPU resources. (Two separate VMs running big.erl get utilization up to 85%.) The answer may be buried somewhere inside operating system limits (see, e.g., sysctl(3) and sysctl(8); maybe kern.clockrate?), but it might also be something more interesting. Meanwhile, I'll try to come up with a similar toy benchmark for Haskell to see if it achieves better utilization of the CPUs.

It turns out that the temperature issues with my G5 were not merely due to a little accumulated dust.  The temperature sensor on CPU B has gone downhill to the point where it puts the machine to sleep after about 10 minutes from a cold boot, whether the machine is just idling or working hard, and then it's another couple of hours before its cool enough to boot again.  The interesting thing to me is that this isn't an isolated issue, see, for example, this post and this thread.  In fact, Apple changed the cooling design after my machine was shipped.

I had a "Genius" at the Apple Store and a local non-Apple repair shop take separate looks at the machine, and the diagnosis was the same — unknown root cause, but the standard course of action is to successively replace the logic board and one or both CPUs to see if that helps.  (I didn't ask the "Genius" to show me his Mensa card...)  The part shuffle isn't a surefire fix: it may help, but it may just cost money and not help at all.  The person-to-person street value for an equivalently configured machine, via eBay, is about $1,800, which is barely more than the total cost of parts and labor if both the logic board and one CPU are needed.  The machine is nice and heavy, so maybe it will make a good boat anchor...

For what it's worth, my experience at the Genius Bar was similar to Paul Boutin's.  A twenty-something woman with various metal bits in her lips and nose asked me what the G5 was and if I did "computer stuff".  I contemplated making up a story about how it was a really big iPod with a special headphone amplifier and personal subwoofer built-in, but I just replied "yes" and left it at that.  The local repair shop provided more information but took two weeks to do it.  Maybe people who use their Apple products to make money instead of spend it are no longer the company's core market?  His house is a lot bigger than mine, so I can hardly fault Real Steve Jobs for his choices.

Today, I spent nearly two hours on the phone with Apple support, working my way through offshore call centers to someone with actual knowledge of the machine, but the response was a resounding "You should have bought AppleCare."  I have a hard time trusting someone who can look me in the eye and tell me that I need to put money down on a bet that their product will fail, but that's the official line.

Assaf put his Twitter finger on one of my thoughts: maybe Apple is not the answer.

And maybe it is time to give Linux another shot as a primary operating system.  The company that I bought reliable workstations from back when I was an academic is still in business...  Apple is far from perfect, but up to now, they have been better enough to win my business.  Like other people, I have had issues with Apple operating system updates on my laptops (SuperDuper is well worth the money!), and I've had to install from scratch at least a half-dozen times in the past five years.  In fact, I've developed a ritual for migrating my settings and additional binaries (primarily MacPorts) to a new install.  Nonetheless, it would be hard to leave the Mac behind.  I use and love OmniGraffle, Keynote, VoodooPad, QuickSilver, TaskPaper, and a host of other Mac-only software, including more spendy stuff like Reaktor.

If the guts are sufficiently generic, I'll give the forthcoming Penryn machines a chance, and if not, I'll take my money elsewhere.

I was actually excited to see iLife '08 come out because of one small feature in iMovie:

I took "AVCHD support" to mean that HD footage from my camcorder (Sony HDR-SR1) was finally supported, but trying to import AVCHD video in iMovie on my G5 pops up an error dialog:

The fine print isn't linked from the iMovie product page or from the iMovie applicaction help, but there is a camcorder support document that links to a knowledgebase article that explains that AVCHD support is Intel only for the HDR-SR1. (Oddly, other cameras appear to have no problems with AVCHD import on the G5?)

That little issue might be worth an asterisk on the feature list.