To a limited extent - yes. There isn't an unlimited amount of things to spend power on.
For example - over the last 5 decades (from memory) the average house temperature in the UK during winter has risen 10C - 18F.
This causes more energy use - however you do it - unless you revamp the housing stock to use less energy. A house built to current specifications consumes less energy to heat than one built 100 years ago, and heated to 1950s standards.
There are two big drivers that might actually cause energy usage to drop - and they are a bit painful.
A) Massive increases in energy costs.
If power was 10* as expensive - people would actually look at stickers when purchasing appliances.
B) Regulation. For example, a (revenue neutral to the government) discount/surcharge on a product, based on the difference of its power usage over similar appliances. TV A uses 1.3* the power of TV B - it costs 30% more - with the difference going to subsidise the maker of TV A.
To have major impacts, you need major changes, which will not happen by fiddling around the edges.
Even - for example - having to include in the sticker price - with the same prominance as the price - the price plus the total energy usage over 5 years would have a large effect.
Anyone that claims 'we are running out of resources' - without specifying 'as we currently use them' - is a complete failure.
Almost nobody at all wants to use more resources. They want certain things, and don't care how they're provided.
For example - lighting. If you take current lighting levels in homes, and compute it out, you end up with the figure that you'd need 15 tons of candles a year to light the average home as well as it now is.
Consider how much it would cost in 1700 to have the countries leading musicians play one 'track' each as background music at a dinner.
Heating/cooling of houses in the best and average homes worldwide is another huge component of energy use that could be improved without anyone caring.
Technology can help enormously with energy use. It's plausible that as LED lighting hits, it's going to reduce energy use of even the best current technology by a factor of 2ish. Aerogel insulation for homes is not intrinsically expensive, and yet could improve dramatically over the normal today, as are many energy saving technologies - air exchange ventilation.
Cars are energy hogs. But even there, it's possible to improve the performance and reduce energy usage - see the various projects in progress to let cars automatically form closely spaced 'road trains' - which will reduce drag.
in short - go and look at a breakdown of resource usage by task, and compare the best plausible or cutting-edge now tech in 20 years, as it could be implemented. There are _huge_ savings to be made.
They really don't. They make very expensive solar cells at the moment - they haven't managed to scale production.
They have funding of over a billion dollars so far. They hope to have production at 115MW/year in 'Fall 2011', with production of 20MW in 2011. (this is on a background of them having announced capacities of around 1GW/year in 2008)
So - 100 dollars a watt or so for produced panels.
Current 'normal' solar panels are down to as low as about a dollar a watt, and falling, making nanosolars claims of $.6/W in 'just a few years' not really look very low cost any more.
I'm not sure that you've not heard confused accounts of bloodlettig, which is used by some tribes to get some nutrition out of the cow without killing it. http://www.youtube.com/watch?v=KRYZp8KmODE Caution - link does not contain Rick Astley.
Being too lazy to read the original paper. Did they attempt to measure calorific expenditure? Calories burned when playing xbox are higher than that when sitting doing nothing.
I'm not quite meaning make it a proper flight recorder, just a basic system.
Acceleration is easy - on a small object, a 2 gram flight recorder has a 'weight' of 7Kg or so. The idea would be that these would be pretty much scattered throughout the structure in known places.
For example - it's rare that none of the engines ever get found, so scatter a dozen in each engine,...
I've wondered for some time why planes don't decentralise this a bit. For example, stick a 1cc little cube with an accelerometer, gyro, and some flash memory into all of the electrically operated emergency lights. All it does issit there and log accelleration, gyro readings, and temperature to flash, and rewrites after a few weeks. This nowadays takes truly modest amounts of power and volume. Engines generally survive - stick a few dozen in there.
That's somewhat variable. Peltiers basically blow if you use them at over half their 'sticker' maximum temperature difference. At that point, they have perhaps a COP of 1. At a temperature difference of around 1/6 maximum - they are up to around a COP of 4-5, which isn't bad at all. However - this is a delta of 6C or so - which isn't really usable in most applications. It's worth noting that a COP of 1 isn't useless. If you can make it cheap enough, you can make a electric heater with double the output.
I'm living off about that much per month in the UK, delivered. And not that unhealthily. Pick the right ingredients, with an eye to cost, proper use of a freezer,...
Indeed - I can order online from any of three supermarkets - Asda, Tesco, Sainsburys,
At the moment, I'm unable to drive, as health problems are making it difficult for me to get my licence. This also means I'm on a severe budget.
Online delivery means I don't have to drag a couple of bags home on the bus every day or three - it's great!
It also means that with the aid of my freezer, I can eat really quite cheaply indeed. I base my orders around buy-one-get-one-free, or half-price offers, and am at the moment shopping around monthly.
The numbers understate military spending. You have to also count 25% of the debt repayment, and the veterans benefits - so it's not 25%, but 30% for the submitter.
Yes. Eventually. Reverse engineering and hacking closed stuff is ____NOT___ a victory. It sends the wrong signals. 'Protected stuff sells just fine'. 'We don't need to worry about little guys stealing our market as the nerds can hack our cheap boxes'. 'Appeasing content providers is an easy buisness model'
The problem with hacking is that it's getting easier to protect stuff. A decade ago, if you were making a router, you had little choice to make it from a CPU chip, a ROM chip, and a RAM chip. All soldered to a board, with comparatively accessible traces.
Ok - worst case, you needed to desolder the flash, and it was really annoying to do.
There is almost no way to protect keys in this beyond the 'normal' code obfuscation methods.
Now, increasingly security architecture is moving on-chip, and becoming cheap. Partially as a side-effect of making devices more flexible. Many or even most small 32 bit chips now have a small area of ROM that handles the initial boot, and some user-settable one-time writable memory.
Because it's 'free' (a K or two), these often now include routines that will let the user on initial flash (or in production of the a large number of chips) say 'only boot from a bootloader with key authenticated by the in-ROM key'
To get to this key is practically very hard - especially if the vendor has taken measures - covering the few bytes of ROM in question with metalisation - to prevent this. You can't get at it with a soldering iron. You can't often now even get at the off-'chip' RAM or ROM easily now, as it's not on seperate chips, it's on chips laminated to the CPU.
Geohot - for example - did nothing at all clever cryptographically. He exploited a basic bug in the implementation that is the sort of thing you get when someone reads the manpage on a crypto function, and implements it, not really understanding all of the twiddly bits, and leaving some out.
Getting crypto right with modern chips is getting increasingly easy - it is not more expensive or needing more hardware to get it right, it simply needs employing someone with a clue to look over your code. Drop 20K on http://www.schneier.com/ - for example - or basically anyone that's actually understood crypto, and is not just writing it as a 'normal' program.
The only 'right' way to respond to this is to buy open platforms. Unfortunately, this is often hard.
"it still remains that a private US business has gone from an initial investment of $100M in 2002 to proposing building a rocket with a considerable fraction of which is already flight proven and twice as much payload as anything else flying today. That's pretty good for less than ten years!"
You want to replace the CPU? Ok - you need a ~400 pin socket, a clip to hold the CPU down, an adaptor to hold that CPU, as it's not designed to be socketed. This adds weight, unreliability, and volume.
You want to add a hole for upgradable wireless - it gets even more complex. You have to pick ahead of time the volume for this hole - which will mean wasting space, as you need to leave some spare. You need an internal lining for this hole. You need a hole in the structure, which weakens it and adds weight and bulk to keep the stiffness up. You need a covering for the 3G add-on, you need a couple of fine-pitch connectors, which raise their own unreliability issues as well as adding weight and volume. You need more decoupling capacitors, as the power supply is now the far side of a high impedence connector. Then you run into the issue that the antenna you have will not be tuned properly.
Sure, this can all be done. However a fully modular phone will cost at least twice, weigh around twice, and be much less reliable.
Miniature connectors are _hellishly_ unreliable. Miniature multiple-hundred-pin ones even more so. I
I am in the fortunate position of owning my own house, and the unfortunate position of having a low income that may be interrupted.
I have around 3 months stock of 'normal' food, partially in a very large freezer. I have maybe another 6 months of 'meh - pancakes again' type food. I do have a generator, but I've chosen to keep a surplus of 6 months electricity paid with my electricity supplier.
Natural disasters are fortunately rare in Scotland. This year I'm insulating the house, from its largely uninsulated prior condition to really quite toasty. This'll mean I can have the heating on more than an electric blanket next winter, which is partially paid for by not having the heating on this winter.
There are a number of common answers to why aliens might have lousy security. For example - if you're generally running about invading/killing civilisations that haven't really gotten beyond cannon for the past few tens of milennia, then external threats may not have been prioritised in your design.
They don't have the source. They bought the OS off an extinct race 100000 years ago, and have just patched it a bit.
Hacking is pointless - if you're a hive species, then there may be little motivation for hacking.
Where it differs is that there is an on-chip regulator to do the dynamic scaling. The TI solution has a couple of regulators on-chip, with a couple of output voltages, as well as a more variable external solution.
The above device has variable regulators on-chip. (for annoying technical reasons, these are linear regulators, not switching, so if they regulate to 50% output - half the (reduced amount of power needed) is wasted as heat.
Diddn't last that long, and was soon left to wither.
The same might happen to valve guitar amps - fashion moves on. The reaction against transistor amps, and reluctance to embrace DSP was largely fueled by some early very bad attempts at it.
The newer generation may not see this as a good reason to continue with valve amps, versus a solid state amp with a 'valve' setting.
Consider for example that many DJs now are moving, or have moved away from vinyl, to various forms of computer based system, even for 'scratching'.
Slashdot Mirror
To a limited extent - yes.
There isn't an unlimited amount of things to spend power on.
For example - over the last 5 decades (from memory) the average house temperature in the UK during winter has risen 10C - 18F.
This causes more energy use - however you do it - unless you revamp the housing stock to use less energy.
A house built to current specifications consumes less energy to heat than one built 100 years ago, and heated to 1950s standards.
There are two big drivers that might actually cause energy usage to drop - and they are a bit painful.
A) Massive increases in energy costs.
If power was 10* as expensive - people would actually look at stickers when purchasing appliances.
B) Regulation.
For example, a (revenue neutral to the government) discount/surcharge on a product, based on the difference of its power usage over similar appliances.
TV A uses 1.3* the power of TV B - it costs 30% more - with the difference going to subsidise the maker of TV A.
To have major impacts, you need major changes, which will not happen by fiddling around the edges.
Even - for example - having to include in the sticker price - with the same prominance as the price - the price plus the total energy usage over 5 years would have a large effect.
Anyone that claims 'we are running out of resources' - without specifying 'as we currently use them' - is a complete failure.
Almost nobody at all wants to use more resources.
They want certain things, and don't care how they're provided.
For example - lighting.
If you take current lighting levels in homes, and compute it out, you end up with the figure that you'd need 15 tons of candles a year to light the average home as well as it now is.
Consider how much it would cost in 1700 to have the countries leading musicians play one 'track' each as background music at a dinner.
Heating/cooling of houses in the best and average homes worldwide is another huge component of energy use that could be improved without anyone caring.
Technology can help enormously with energy use.
It's plausible that as LED lighting hits, it's going to reduce energy use of even the best current technology by a factor of 2ish.
Aerogel insulation for homes is not intrinsically expensive, and yet could improve dramatically over the normal today, as are many energy saving technologies - air exchange ventilation.
Cars are energy hogs. But even there, it's possible to improve the performance and reduce energy usage - see the various projects in progress to let cars automatically form closely spaced 'road trains' - which will reduce drag.
in short - go and look at a breakdown of resource usage by task, and compare the best plausible or cutting-edge now tech in 20 years, as it could be implemented.
There are _huge_ savings to be made.
They really don't.
They make very expensive solar cells at the moment - they haven't managed to scale production.
They have funding of over a billion dollars so far.
They hope to have production at 115MW/year in 'Fall 2011', with production of 20MW in 2011.
(this is on a background of them having announced capacities of around 1GW/year in 2008)
So - 100 dollars a watt or so for produced panels.
Current 'normal' solar panels are down to as low as about a dollar a watt, and falling, making nanosolars claims of $.6/W in 'just a few years' not really look very low cost any more.
I'm not sure that you've not heard confused accounts of bloodlettig, which is used by some tribes to get some nutrition out of the cow without killing it.
http://www.youtube.com/watch?v=KRYZp8KmODE
Caution - link does not contain Rick Astley.
I F***N wish.
I have a relatively uninsulated 100m^2 home.
1.6Kw raises the internal temperature by around 3C or so.
Indeed.
Some 'similar' activity would be an appropriate control.
Perhaps knitting, or building a lego fort.
Being too lazy to read the original paper.
Did they attempt to measure calorific expenditure?
Calories burned when playing xbox are higher than that when sitting doing nothing.
I'm not quite meaning make it a proper flight recorder, just a basic system.
Acceleration is easy - on a small object, a 2 gram flight recorder has a 'weight' of 7Kg or so.
The idea would be that these would be pretty much scattered throughout the structure in known places.
For example - it's rare that none of the engines ever get found, so scatter a dozen in each engine, ...
I've wondered for some time why planes don't decentralise this a bit.
For example, stick a 1cc little cube with an accelerometer, gyro, and some flash memory into all of the electrically operated emergency lights.
All it does issit there and log accelleration, gyro readings, and temperature to flash, and rewrites after a few weeks.
This nowadays takes truly modest amounts of power and volume.
Engines generally survive - stick a few dozen in there.
Once they are - say - 10 years old - the complete source and schematics of the ECU, as well as all other parts of the car are revealed.
That's somewhat variable.
Peltiers basically blow if you use them at over half their 'sticker' maximum temperature difference.
At that point, they have perhaps a COP of 1.
At a temperature difference of around 1/6 maximum - they are up to around a COP of 4-5, which
isn't bad at all.
However - this is a delta of 6C or so - which isn't really usable in most applications.
It's worth noting that a COP of 1 isn't useless.
If you can make it cheap enough, you can make a electric heater with double the output.
I should have actually saved the graph I made.
To replicate - go to http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=102-1663-ND get the datasheet - now refactor the graph into power in vs heat pump capacity over temperature and current.
I'm living off about that much per month in the UK, delivered. ...
And not that unhealthily.
Pick the right ingredients, with an eye to cost, proper use of a freezer,
Indeed - I can order online from any of three supermarkets - Asda, Tesco, Sainsburys,
At the moment, I'm unable to drive, as health problems are making it difficult for me
to get my licence. This also means I'm on a severe budget.
Online delivery means I don't have to drag a couple of bags home on the bus every
day or three - it's great!
It also means that with the aid of my freezer, I can eat really quite cheaply indeed.
I base my orders around buy-one-get-one-free, or half-price offers, and am at the
moment shopping around monthly.
One of them even has an online API! http://www.techfortesco.com/forum/index.php?board=1.0
Being able to complete an order at leisure, and to reflect on each purchases value and
calories/... has greatly trimmed my grocery bill and waistline.
I'm in a small village - 6 miles from the nearest town of 40K - no 'fast food' places will deliver.
The numbers understate military spending.
You have to also count 25% of the debt repayment, and the veterans benefits - so it's not 25%, but 30% for the submitter.
You can't microprobe under metalisation.
Yes. Eventually.
Reverse engineering and hacking closed stuff is ____NOT___ a victory.
It sends the wrong signals.
'Protected stuff sells just fine'.
'We don't need to worry about little guys stealing our market as the nerds can hack our cheap boxes'.
'Appeasing content providers is an easy buisness model'
The problem with hacking is that it's getting easier to protect stuff.
A decade ago, if you were making a router, you had little choice to make it from a CPU chip, a ROM chip, and a RAM chip.
All soldered to a board, with comparatively accessible traces.
Ok - worst case, you needed to desolder the flash, and it was really annoying to do.
There is almost no way to protect keys in this beyond the 'normal' code obfuscation methods.
Now, increasingly security architecture is moving on-chip, and becoming cheap. Partially as a
side-effect of making devices more flexible.
Many or even most small 32 bit chips now have a small area of ROM that handles the initial boot,
and some user-settable one-time writable memory.
Because it's 'free' (a K or two), these often now include routines that will let the user on initial flash
(or in production of the a large number of chips) say 'only boot from a bootloader with key authenticated
by the in-ROM key'
To get to this key is practically very hard - especially if the vendor has taken measures - covering the few
bytes of ROM in question with metalisation - to prevent this.
You can't get at it with a soldering iron.
You can't often now even get at the off-'chip' RAM or ROM easily now, as it's not on seperate chips, it's on
chips laminated to the CPU.
Geohot - for example - did nothing at all clever cryptographically.
He exploited a basic bug in the implementation that is the sort of thing you get when someone reads the
manpage on a crypto function, and implements it, not really understanding all of the twiddly bits, and leaving
some out.
Getting crypto right with modern chips is getting increasingly easy - it is not more expensive or needing more
hardware to get it right, it simply needs employing someone with a clue to look over your code.
Drop 20K on http://www.schneier.com/ - for example - or basically anyone that's actually understood crypto,
and is not just writing it as a 'normal' program.
The only 'right' way to respond to this is to buy open platforms.
Unfortunately, this is often hard.
"it still remains that a private US business has gone from an initial investment of $100M in 2002 to proposing building a rocket with a considerable fraction of which is already flight proven and twice as much payload as anything else flying today. That's pretty good for less than ten years!"
$100M is hardly nothing.
It's a magnificent achievement, yes.
Which means little - you can get twice the nominal range out of a Prius, using hypermiling techniques.
Modularity has costs.
You want to replace the CPU?
Ok - you need a ~400 pin socket, a clip to hold the CPU down, an adaptor to hold that CPU, as it's not designed to be socketed.
This adds weight, unreliability, and volume.
You want to add a hole for upgradable wireless - it gets even more complex.
You have to pick ahead of time the volume for this hole - which will mean wasting space, as you need to leave some spare.
You need an internal lining for this hole. You need a hole in the structure, which weakens it and adds weight and bulk to keep the stiffness up.
You need a covering for the 3G add-on, you need a couple of fine-pitch connectors, which raise their own unreliability issues as well as adding weight and volume.
You need more decoupling capacitors, as the power supply is now the far side of a high impedence connector.
Then you run into the issue that the antenna you have will not be tuned properly.
Sure, this can all be done. However a fully modular phone will cost at least twice, weigh around twice, and be much less reliable.
Miniature connectors are _hellishly_ unreliable.
Miniature multiple-hundred-pin ones even more so.
I
They must be really scary then.
I am in the fortunate position of owning my own house, and the unfortunate position of having a low income that may be interrupted.
I have around 3 months stock of 'normal' food, partially in a very large freezer.
I have maybe another 6 months of 'meh - pancakes again' type food.
I do have a generator, but I've chosen to keep a surplus of 6 months electricity paid with my electricity supplier.
Natural disasters are fortunately rare in Scotland.
This year I'm insulating the house, from its largely uninsulated prior condition to really quite toasty.
This'll mean I can have the heating on more than an electric blanket next winter, which is partially paid for
by not having the heating on this winter.
Halle Berry, with a gun, and not much else. Makes up for a hell of a lot of plot holes.
There are a number of common answers to why aliens might have lousy security.
For example - if you're generally running about invading/killing civilisations that haven't really gotten beyond cannon for the past few tens of milennia, then external threats may not have been prioritised in your design.
They don't have the source. They bought the OS off an extinct race 100000 years ago, and have just patched it a bit.
Hacking is pointless - if you're a hive species, then there may be little motivation for hacking.
Sounds similar to SmartReflex (tm) which is shipping on millions of phones.
http://focus.ti.com/general/docs/wtbu/wtbugencontent.tsp?templateId=6123&navigationId=12032&contentId=4609&DCMP=WTBU&HQS=ProductBulletin+PR+smartreflex
Where it differs is that there is an on-chip regulator to do the dynamic scaling.
The TI solution has a couple of regulators on-chip, with a couple of output voltages, as well as a more variable external solution.
The above device has variable regulators on-chip. (for annoying technical reasons, these are linear regulators, not switching,
so if they regulate to 50% output - half the (reduced amount of power needed) is wasted as heat.
I question this somewhat.
Fashion in music is hard to predict.
http://en.wikipedia.org/wiki/Skiffle - for example.
A retro take on 1930s music, in the 1950s.
Diddn't last that long, and was soon left to wither.
The same might happen to valve guitar amps - fashion moves on.
The reaction against transistor amps, and reluctance to embrace DSP was largely fueled by some early very bad attempts at it.
The newer generation may not see this as a good reason to continue with valve amps, versus a solid state amp with a 'valve' setting.
Consider for example that many DJs now are moving, or have moved away from vinyl, to various forms of computer based system, even for 'scratching'.