Japan Wind Development has opened a 51 MW wind farm that incorporates a 34 MW sodium sulfur battery system at Futamata in Aomori Prefecture in May 2008.[7][9]
I mean "past its peak" in the sense that the dot com bubble was "past its peak" in 1999, even though the literal peak was in 2000. The last bit was just inertia.
And I'm not talking about extinction, just the loss of radio technology for instance. Something that would make us completely undetectable as an intelligent life form viewed by somebody living a few light years away.
Sure, but is it going to be ready in time ? Running cars on nukes will require lots of new reactors, a complete overhaul of the grid, better battery technology, and replacement of about a billion cars. Even if we knew what to do, how long would the implementation take ? Note that the nuclear option isn't very popular right now.
That's why the IPCC has a couple of different scenarios on how the CO2 levels will go. Based on these CO2 scenarios, they have developed temperature calculations.
Of course, nobody can predict the exact CO2 scenario we're going to follow, but despite that, it still makes sense to project a couple of possible paths.
They're not talking about an increase in measured atmospheric CO2. This is an estimate from the International Energy Agency, based on amounts of fossil fuels burned.
Other sources of CO2 will have to be added to the IEA number.
The effects don't enter the oceans immediately. It will take decades for the large volume of water in our oceans to reach a new equilibrium. In the meantime, the atmosphere will keep getting warmer, even if CO2 remains constant for that time.
(eg if cloudcover increases, the earth will be cooler)
Cloud cover also blocks heat escaping at night. The net effect is pretty small. Extra water vapor in the air (which is already happening) does have a warming effect, as it only blocks IR, and not visible light.
2. Every single intelligent civilization self destructs. This also seems ludicrous...even if it happens some times, there should at least be remnants.
We've had already two dozen civilizations on earth that self-destructed, so this seems like a likely scenario. The remnants are likely too hard to detect. Our current civilization is pretty much undetectable beyond the orbit of Pluto, and is probably already past its peak.
A better solution would be to extract the other 99% of the energy still locked in the "waste". Permanently disposing of such a valuable asset, and then discovering the perfect way to use it 10 years later would be mildly annoying.
Consciousness is not a black/white thing. It's more of a continuous gliding scale. Fruit flies have very little, and we have a lot, but there's no magic threshold.
Randomness doesn't add anything beneficial to free will. If I'm about to cross a street, and suddenly decide not too, because I see that the approaching car is going pretty fast, I don't need a random input that says: "you know what, just cross anyway".
A deterministic free will is actually better for survival. Our brains have evolved to exclude randomness as much as possible.
It's not necessary for doctors. To be a successful programmer you need a certain way of algorithmic thinking that takes a certain talent to master. It's about truly understanding something, and applying it in new situations.
To be a successful doctor, you need to be able to memorize lots of things, but there's nothing to 'understand'.
Shouldn't a 32-bit CPU handle code and data accesses better than an 8-bit CPU? I'm not really a CPU architecture guy, so please correct me if I'm wrong.
He means that ARM SoC usually have some jitter in the instruction timing This could be caused by cache fills, or by more complex memory buses, such as a bus which is also shared by a separate DMA controller. Because of the higher bus speeds, there is usually more delay, and this means longer burst access. So the DMA could be doing a long burst, while the CPU has to wait. Also, the Flash may not be running at 72 MHz, but maybe at 18 MHz, and have a 128 bit wide interface + a small cache.
In practice, this is only a problem in the most demanding applications, where you cannot tolerate sub-microsecond jitter. For those applications, it is unlikely that the slow AVR is better suited.
If all you want is a blinking LED, you don't need an ARM, but if you want to have a blinking LED controlled by a built-in web server, an ARM will work nicely.
And why worry about some timing jitter when you're blinking a LED ? What kind of project are you doing that a few dozen nanoseconds is going to matter ?
Having written plenty of code for both ARM and AVR, I have to disagree here. The ARM instruction set is very nice, powerful, and easy to learn. In comparison, the AVR isn't bad, but I certainly wouldn't favor it. The biggest problem with the AVR is the strict Harvard architecture, which means there's no good way that you can make a routine on the AVR that accepts either a pointer to RAM (such as a variable string) or pointer to Flash (such as a constant string).
Besides, you can write many interesting programs without ever having to know the instruction set or the processor model.
Slashdot Mirror
http://en.wikipedia.org/wiki/Sodium-sulfur_battery
That average energy consumption isn't just electricity. Average electricity per person is just 1460 W for the US, which is what this system is for.
I don't know where you live, but we've had double meters (night and day) for as long as I can remember.
So, you're saying we could get a sunburn by staying out too long at night ?
Power levels of bluetooth are 100x lower.
True, but it doesn't say anything about intelligent or complex life.
I mean "past its peak" in the sense that the dot com bubble was "past its peak" in 1999, even though the literal peak was in 2000. The last bit was just inertia.
And I'm not talking about extinction, just the loss of radio technology for instance. Something that would make us completely undetectable as an intelligent life form viewed by somebody living a few light years away.
Sure, but is it going to be ready in time ? Running cars on nukes will require lots of new reactors, a complete overhaul of the grid, better battery technology, and replacement of about a billion cars. Even if we knew what to do, how long would the implementation take ? Note that the nuclear option isn't very popular right now.
We're running out of cheap oil, and have nothing planned for its replacement that can be ready in time. Not even the Chinese.
That's why the IPCC has a couple of different scenarios on how the CO2 levels will go. Based on these CO2 scenarios, they have developed temperature calculations.
Of course, nobody can predict the exact CO2 scenario we're going to follow, but despite that, it still makes sense to project a couple of possible paths.
Especially if it's soft and gooey.
And how many of those civilizations in our galaxy are close enough that we actually have a decent chance of picking up on their communications ?
Unless somebody is aiming a high power directional antenna right at the earth, the signal is going to drown in background noise very quickly.
You're forgetting the Bs factor:
http://xkcd.com/384/
Actually, the Moon's influence on the tides is only 2.21 times larger than the Sun's:
http://imagine.gsfc.nasa.gov/docs/ask_astro/answers/961029b.html
They're not talking about an increase in measured atmospheric CO2. This is an estimate from the International Energy Agency, based on amounts of fossil fuels burned.
Other sources of CO2 will have to be added to the IEA number.
We've stopped coming out of the last ice age about 8,000 years ago.
http://en.wikipedia.org/wiki/File:Holocene_Temperature_Variations.png
The effects don't enter the oceans immediately. It will take decades for the large volume of water in our oceans to reach a new equilibrium. In the meantime, the atmosphere will keep getting warmer, even if CO2 remains constant for that time.
Cloud cover also blocks heat escaping at night. The net effect is pretty small. Extra water vapor in the air (which is already happening) does have a warming effect, as it only blocks IR, and not visible light.
We've had already two dozen civilizations on earth that self-destructed, so this seems like a likely scenario. The remnants are likely too hard to detect. Our current civilization is pretty much undetectable beyond the orbit of Pluto, and is probably already past its peak.
A better solution would be to extract the other 99% of the energy still locked in the "waste". Permanently disposing of such a valuable asset, and then discovering the perfect way to use it 10 years later would be mildly annoying.
Consciousness is not a black/white thing. It's more of a continuous gliding scale. Fruit flies have very little, and we have a lot, but there's no magic threshold.
Randomness doesn't add anything beneficial to free will. If I'm about to cross a street, and suddenly decide not too, because I see that the approaching car is going pretty fast, I don't need a random input that says: "you know what, just cross anyway".
A deterministic free will is actually better for survival. Our brains have evolved to exclude randomness as much as possible.
It's not necessary for doctors. To be a successful programmer you need a certain way of algorithmic thinking that takes a certain talent to master. It's about truly understanding something, and applying it in new situations.
To be a successful doctor, you need to be able to memorize lots of things, but there's nothing to 'understand'.
He means that ARM SoC usually have some jitter in the instruction timing This could be caused by cache fills, or by more complex memory buses, such as a bus which is also shared by a separate DMA controller. Because of the higher bus speeds, there is usually more delay, and this means longer burst access. So the DMA could be doing a long burst, while the CPU has to wait. Also, the Flash may not be running at 72 MHz, but maybe at 18 MHz, and have a 128 bit wide interface + a small cache.
In practice, this is only a problem in the most demanding applications, where you cannot tolerate sub-microsecond jitter. For those applications, it is unlikely that the slow AVR is better suited.
If all you want is a blinking LED, you don't need an ARM, but if you want to have a blinking LED controlled by a built-in web server, an ARM will work nicely.
And why worry about some timing jitter when you're blinking a LED ? What kind of project are you doing that a few dozen nanoseconds is going to matter ?
Having written plenty of code for both ARM and AVR, I have to disagree here. The ARM instruction set is very nice, powerful, and easy to learn. In comparison, the AVR isn't bad, but I certainly wouldn't favor it. The biggest problem with the AVR is the strict Harvard architecture, which means there's no good way that you can make a routine on the AVR that accepts either a pointer to RAM (such as a variable string) or pointer to Flash (such as a constant string).
Besides, you can write many interesting programs without ever having to know the instruction set or the processor model.