28 June 2007

Server upgrade imminent. I have purchased a new motherboard, CPU and RAM to replace the aging Pentium 3 + OR840 combo with 128MB RamBus RAM I am currently using in the server. The new motherboard is a SuperMicro PDSGE and will have a Pentium 4 650 and 1Gb of RAM. I am probably not going to use the new motherboard's on-board gigabit ethernet adaptor because of concerns about Intel's AMT technologies.... Heck, I may just place the AMD PC/Net PCI adaptor for the outward facing interface (Yes, that is the same adaptor I 'cloned' when I wrote my AMD PC/Net PCI emulation for QEmu which is incidentially also used by Xen when performing machine emulation). I may just use the on-board gigabit NIC for a direct crossover link to the Mac for fast jumbo frames. Meanwhile, I just have to make sure that I have everything - Likely to need a new PSU to power the motherboard. Will probably need a few new case fan, good practice to replace them anyways for a machine which practically stays on 24 hours a day.

13 June 2007

Every now and again, I have to have a large sigh...


*** sighs ***


With the recent release from Perdue detailing a computer simulation of the 9/11 WTC attack, the fierce debates that it spawns as to the "It's an inside job" vs. "No it's not," camps and their arguments miss the point of the simulation.
I have looked at the short video clip and I would say that it seems plausible: I could imagine a aeroplane causing that kind of damage. It is also clear to me that the damage rendered to the core columns is inadequate to cause a complete structural failure of the whole building, especially considering that the aeroplanes hit quite high up on the building. There was relatively little load being supported by the core columns at that level.
Of course, the "Terrorists hate us" camp would trot out the argument that steel only needs to be heated up to half it's melting temperature in order to begin to warp and fail. Given that the buildings were hit some 400m above ground level, assuming that there was enough fuel burning to heat the core columns via radiation, and assuming that 100% of the radiation was absorbed with no radiative losses... for that heat to propagate down the core columns such that all the columns were heated sufficiently that a cascade "pancake" collapse would occur, hmm... after some scribblings on a back of a napkin.... assuming a rather light core column at an average cross section of .34 sq m per column... it would take as much energy as released in 18 million pounds of TNT to raise all the core columns to 800 degrees Celsius... Or expressed as a speed of a 747 aeroplane if translated into kinetic energy - about the same as if it was traveling at around 10,000 miles per hour.

Pretty fast, innit?

Oh well... People never let facts bother their perception of the world. Unfortunately, they let any idiot vote nowadays.