I'd like to upgrade to a dual core Opteron but they're damned expensive. However I think the basic premise is correct that Moore's Law is finished. The problem boils down to heat. Modern CPU chips pump out an amazing amount of heat.

Here's a tale of stupidity: A couple years ago I bought a SMP mother board for two dual Athlon-XPs. There's a hack one can do to an XP that converts it into an MP. However once you do the hack, you can't send the CPU back for a refund. How do I know the new CPUs will actually bootup? I have to first install the CPU, put on the thermal grease and then install the cooling fans. If the CPUs boot, then I take the whole works apart, do the hack and then reassemble everything. Anyway, getting on the CPU fan was really a bitch. The damned cooling fan spring simply did not want to mate onto the CPU's socket. As I was using pliers in a futile attempt to get this spring on, I was thinking to myself: "Murphy's Law will cause the pliers to slip, scratch the motherboard and I'll have an expensive motherboard that won't boot". So I then had my moment of brilliance: I'll simply set the coolers on top of the CPUs and not lock them down with the spring (there's still an air gap between the cooler and the CPU chip). I'll then turn on the machine just long enough to see it boot and then quickly turn the machine off. So I do this bit of brilliance and there's instantly this strong smell of burning electronics. The computer flat out will not boot. Now after the fact, I down load from AMD's website the spec. sheet for the XP-1700. The thermal power of the chip is 44.9 watts. However the die size of the chip is about 1 square centimeter. Therefore the heatflux is 44.9 watts/cm^2. The peak stagnation point heat flux of the Mars Exploration Rover aeroshell was 39.64 watts/cm^2. Arrgh!! I'm a buffoon! The chip was hotter than MER's aeroshell!! No wonder it burned up instantly.

Anyway, given that a CPU chip's continuous operating heat flux is greater than the peak heat flux of a spacecraft entering an atmosphere, it's no surprise that Moore's Law has reached its limit.

Gary