I was recently at a presentation where one of the speakers demonstrated some Siemens PLM software - high detail 3D CAD models, stuff like that. One of the benefits they pointed out was that it ran well even on the demonstrator's laptop computer - you didn't need a big workstation to run it on.

However, the demonstrator's laptop was a huge 17" desktop replacement thing.

That's the market I see for these things - mobile workstations that let you run workstation software at client sites and trade show booths, without needing to muck around with carrying and powering a second monitor along with your laptop.

Furthermore, computer companies derive some benefit from holding the "performance crown" even if it's with a product hardly anyone would buy. For example, you can load your computer with two $500 graphics cards or an "extreme" $1600 CPU but probably few people do. I think the idea is to get news coverage for your product line and demonstrate how cutting edge your product line is.

Or, given that we're comparing price and performance, a scatter plot.

I decided to replot some of the graphs properly. Here are the results.

Those are some impressive technologies, but the poster above was talking about charging from home electrical sockets - where the limiting factor isn't the battery but the power supply.

Let's say you wanted to make a 100 mile round trip, driving for 2 hours at 50mph. Wikipedia tells me a car cruising at 50mph needs about 10 horsepower, or 7.5kw, to overcome drag. So, 7.5kw for 2 hours.

A socket providing 110 volts, 15 amps gives 1650 watts - so charging would take 2*7500/1650 = 9 hours. If that was a 240 volt 13 amp socket (like in the UK) you get 3120 watts - charge in 4.8 hours. The power switchboard for my house can take at most 100 amps at 240 volts - 24,000 watts, for a charging time of 37 minutes. A typical industrial socket might give 400 volts and 40 amps on each of 3 phases. That's 48,000 watts and you could charge in 18 minutes.

And that's one reason people talk about special electric car power infrastructure: A 400 volt 3 phase 40 amp supply to every house would need a major infrastructure upgrade. You can do clever things like only charging during periods of high supply/low demand, but that requires control infrastructure too.

And of course, the figures given above are for a 10 horse power car. And let's be honest here - even small cars like the Ford Ka have 50 horsepower engines.

Well, a community is one way to prevent spamming, but is it the only way?

I mean, Wikipedia goes for a very transparent structure. If Wikipedia added secret algorithms to selectively display only certain users' contributions, they would probably be criticised by people on Slashdot, and elements of the Wikipedia community.

On the other hand, Google already uses secret automated moderation techniques for search results, and they are also pretty good at filtering spam in gmail. Both with no need for a community of volunteers doing filtering manually.

Furthermore, Google has not committed to using this user-generated data at all, let alone in an unmoderated form.

Now, I agree with you that Google hasn't got a track record of building user communities like Wikipedia. However, I think they could make user-moderated search work without one.

The end user retains the freedom to use the BSD code. The fact that someone has released a proprietary product based on the BSD code does not diminish the BSD code.

I think you've just answered your own question by saying "What do people want additional CPU power for?" and saying "I'd rather have a firewire port to connect my video camera to".

I can see AMD's point, though; it does seem that, when ordering laptops by price, screen size goes 9", then 10", then 15", then back to 13".