First off, the batteries are not "laptop" batteries per se. They are 18650 lithium ion cells, like these. While this ad shows them as "3.7 volts" and "1300-1400 mAh," the nominal voltage is considered 3.6 volts; let's split the difference and call them 1350 mAh. That means (3.6v x 1.350 Ah = ) 4.86 Wh / battery. A two-pack is $12.99, which mean ($12.99 / 9.72 Wh = ) $1.34 / Wh. At that rate, 50 kWh (which the article states is the car's capacity) costs about $67,000.
Their tZero gets at least 5 miles / kWh (according to other sources), which means at least 250 miles / charge (the EV1 got closer to 6 miles / kWh).
Note: many laptop battery packs (especially the third-party, aftermarket ones) are simply plastic enclosures with these replacement Li-Ion batteries in them. Hence the comment about "laptop batteries."
Still too pricey for making my own Battery Electric Vehicle (the batteries alone are more than I spent on my last three cars, combined), but we're DEFINITELY getting there. A year ago, the best deal I could find was over $11 / Wh.
If you've got RealPlayer installed on your machine, EVWorld.com has an interview with one of the car's builders, providing some more technical detail.
I mean, the basic tenet is that if you use more gasoline, you will pay more money in taxes. And yes, under the current system, most of the money collected on those taxes is supposed to go to road repairs.
Take a look at this article. It points out that about 23% of the price of gasoline in the US is tax. In the UK, it's 76%. Any question about why petrol (gasoline) works out to about $4.37/gallon
Read an article or two about people in the UK complaining about how much it costs to drive anywhere, and think again.
A couple comments on yours. You make some excellent points.
NREL typically rates "one sun" of solar flux as approx. 1 kW / square meter. Typical solar panels are, as you state, about 15% efficient. If you want an idea of how much solar flux your area gets, take a look at the maps from these pages.
The best hydrolyzers (electrolysis units) are about 60% efficient. Since a kilogram of hydrogen contains about 33 kWh of energy (if you convert at 100% efficiency), a 60% efficient hydrolyzer will need about 55 kWh to make one kilogram of hydrogen. A kilo of hydrogen has roughly the same energy content as a gallon of gasoline. Consequently, if you're going to burn hydrogen in your ICE vehicle, you'll need 55 kWh of energy to replace one gallon of gasoline. In case you're interested, that's about 14 LITERS of liquid hydrogen (density = 71 grams / liter); no other commercially available hydrogen storage can match LH2 for energy density (yeah, there are some experimental systems which have been announced, but none of the COMMERCIALLY AVAILABLE solutions are anywhere close). Go find seven 2-liter bottles of soda (or which used to contain soda), and consider hauling that volume around to replace 3.785 liters of gasoline. Any question why they're having such a hard time storing a significant amount of hydrogen?
Last, but not least, most of your mobile fuel cell stacks are only 40% efficient. That beats the ICE's by at least 50% (last time I checked, none of the ICE's are beating 25% efficiency). Let's see, 60% effeciency at the hydrolyzer, and 40% efficiency at the fuel cell equals (0.6 x 0.4 = 0.24) 24% of the electricity you fed into the hydrolyzer actually coming out of the fuel cell. Talk about wasteful.
By the time you consider how little fuel you'll be able to store, you might as well just build yourself a battery-driven electric vehicle. The GM Impact/EV II gets about 6 miles / kWh. It has a range of about 96 miles (works out to 16 kWh storage). Over 75% of the energy fed to it comes back out; beats the hell out of 24%. Also, considering the price you'd pay for the fuel cell stack (Toyota's Fuel Cell Vehicle is quoted as costing $250,000 per unit to build; they lease them, not sell them), you could probably buy NIMH or Li-Ion batteries and get some pretty impressive range.
Any argument for a fuel cell vehicle is a bigger argument for battery electric vehicles. Safety issues aside.
O.k. Where's the "gotcha" clause? I have a hard time believing that MS is actually going open-source on ANY of their IP. After all it's their "Intellectual Property," right? And if it's "property," sharing it or giving it away would reduce share value, right?
You (and various other posters on this topic) have pointed out that an FPGA provides its best advantage on hardware-accelerating certain things. Other posters have pointed out that we aren't looking at true reconfigurable computing anywhere in the next few years, as actually writing calculations to hardware is very tedious.
O.k., so let's meet somewhere in the middle. Another poster pointed out that a CPU's ISA can be thought of as an API. So, let's build an expanded API. More specifically:
create a library of standard routines which can benefit from hardware acceleration
build solid HDL descriptions of these routines
add an FPGA to your system architecture (add-on card for existing systems, built-in on newer motherboards)
put together something which can load/unload routines on said FPGA
enjoy hardware-accelerated performance benefits
There's probably a:
4. Profit!
in there somewhere.
Think about it folks; how many of us have.so or.dll files on our machines (I'm guessing the answer is ALL of us)? Software interfaces with those libraries, treating them like a sort of black box where parameters go in and results come out, and the programmer can get on with writing the app, rather than re-inventing someone else's work. Think of this thing as a "hardware.dll" (or "hardware.so" for us Unix lovers). With a software library, you have to load the desired routine(s) into RAM (usually handled transparently), then use them. Well, you'd have to load the desired routine into the FPGA, then have a software wrapper which takes library calls and re-routes them to the FPGA, sending the results back. I can even envision a situation where you have a software library AND a matching FPGA library, and you can configure your system to run that library on hardware for acceleration, and the apps which use the library don't have to do any thing different.
Just how much of libc6 would benefit from FPGA acceleration?;-) Which.dll/.so is used the most in Quake/UT/gory-shoot-em-up-of-the-minute?
Find his book, "The Hydrogen World View," by Roger E. Billings (ISBN 0963163426). It's autobiographical, an amusing read, and packed with A GREAT DEAL of valuable insight.
Pegasus, gasogen, wood gas, call it whatever you want. It's all carbon monoxide, and the process is called partial oxidation.
Simple fact: heat some carbon-based material to its burning temperature in an environment which doesn't have enough oxygen. It will produce carbon monoxide (if there's enough O2, you will get CO2). Two CO molecules + one O2 molecule = CO2 + heat.
Yes, this was used heavily in France during WWII. The Germans were taking all the petrol they could get their hands on to fuel their equipment, so the civilians had to be "clever" if they wanted to still use their vehicles. Additionally, back in the '70's, some guys (Americans; saw this on Real People, if you can believe it) put a device on the back end of a small station wagon, fed it with wood scraps from lumber yards and sawmills, and drove cross country. They didn't have to pay for their fuel; most places were happy to let them take all the "waste" they wanted. Part of the wood was spent heating the oxidation chamber, part of it was going into the oxidation chamber. The resulting CO was piped under the car to the engine (a six-cylinder, IIRC). It wasn't the fastest thing on the road (you had limits to how fast you could feed it CO, so the power was kinda wimpy), but it did get out on the interstate and keep up. Considering the fact that the rest of the economy was struggling with the "energy crisis" at the time, they were trying to raise awareness to the fact that other solutions existed.
Considering the fact that Nazis were fond of using CO to kill large numbers of Jews during the Holocaust, I'd say play with this tech at your own risk.
It's a known fact that you can run an Internal Combustion Engine on hydrogen gas.
The person to go looking for on this one is Roger Billings; he's currently with the International Academy of Science (I drive by there every morning on my way to work). He was driving hydrogen-powered ICE vehicles back in the '70's.
He ran into two problems with running an engine on Hydrogen.
Backfiring; he details how, on his first attempt to power a lawn-mower engine on hydrogen gas (he was in his teens) he nearly blew himself and his brother up when the engine backfired. He later overcame the problem (years later) with a water injection system which kept the intake air cool enough that the ultra-volatile hydrogen wouldn't ignite before the spark plug fired
NOx emissions; basically, when the temperature in the combustion chamber gets above 700F, you get more Nitric Oxides building up (NItrogen and Oxygen in the air bonding together), and hydrogen can EASILY beat that temperature. Again, the water injection system kept the max temp below the level, keeping the NOx emissions down.
He got hydrogen engines down to such an artform that he modified a Volkswagen Beetle ('72, IIRC) to run on the stuff for a college competition (he was an undergrad at the time), and the emissions coming out of the tailpipe were actually CLEANER than the air going into the intake. Basically, any Carbon Monoxide or unburned Hydrocarbons (common vehicle pollutants) which made it into the intake were finished off in the process, and the hydrogen fuel didn't produce any such emissions (water vapor).
I'd be wary about adding hydrogen to a gasoline engine to help the economy. You'd need a significant amount of hydrogen to make any real difference, and hydrogen storage these days is either:
high-pressure gas (expensive, heavy tanks and very little capacity)
liquid hydrogen (cryogenic materials, with associated hazards)
metal hydride (the safest of the bunch, but still very expensive for significant amounts of storage)
Want a solution to the problem? It's called a plug-in hybrid. It runs about 20-40 miles on a charge, then kicks in an ICE to provide power after that. Result: most of the small "grocery getter" trips are done pure EV, but you can still fill up at any gas station if you go out of range.
The major auto manufacturers aren't doing this yet, but some colleges are.
Agreed. Someone mentioned a page with benchmarks of an 800 MHz, but that was an Ezra-core. The new Nehemiah core uses DDR RAM (which helps TREMENDOUSLY with the video bandwidth; the CPU doesn't use all the bandwidth, so the video uses the rest of it).
Supposedly, the newer Nehemiah-core systems are capable of playing MPEG-2 and DivX full-rate, which the older systems simply couldn't do.
This tech is much older. Take a look at this article (note: it's a.pdf file). I first read about this stuff in 1993. Texas Instruments started developing this 1983 (yes, that's two decades ago), finally abandoned it and licensed it to someone else.
Hydrazine is indeed nasty stuff. They also use it on Emergency Power Units (EPU's) on the F-16. The training they provided (for those of us who worked on the F-16) indicated that, if exposed to it, your chance of developing leukemia increases about 1,000x. Not to mention the fact that it tends to burn at about 1,600 F and is EXTREMELY caustic. We lost an F-16 (an older one; it was in the local reserve unit) while I was stationed in Utah, and the news reports REPEATEDLY told people to stay the hell away from the wreckage, due to some of the noxious chemicals on board (jet fuel isn't exactly nice stuff, but it's nothing compared to hydrazine).
Hydrazine is commonly used for such backup power systems, including on the shuttle, because it is very compact and very simple to design in. Yes, if it leaks you need people in the full chem/bio suits to clean it up. But, if properly stored and used, it is very handy stuff.
I don't know about the eco-system part, but the phase-change aspect has me wondering.
According to the article, the external temperatures on the CPU cooler box was 24 C. According to this page, if you reduce the pressure in your water system down to 3% of 1 atm pressure (relative vacuum), that would be enough to get the water to boil. Once you got the pressure down, you probably wouldn't need any kind of pump. Simply have the water boil at the CPU cooler, then run it through a long heat-sink attached to a copper tube. The water would cool enough to condense back to a fluid, running back down the tube, and repeating the cycle.
Oh, yeah. That already exists; it's called a heat pipe. The one in the linked article is relatively small and has a fan blowing through a radiator for active cooling. How long until somebody out there "rolls their own," with a long, passive heatsink? The author of the article was looking for "quiet," and that would certainly fill the bill.
Yeah, imagine a beowulf cluster of them. When I was in college, a friend of mine was building a beowulf cluster of 386 machines. His main workstation was a 486 (note: the campus computer lab had year-old Pentium 233 machines). His entire cluster was about as fast as a machine in the campus computer lab.
Why did he do this? We wanted some experience with building, managing and programming a cluster. It cost him less than $200 to build a 10 machine cluster, and he could honestly say he had some experience with it. When he wanted to get something done quickly, however, he walked to the campus computer lab and used one of their machines. Unless you're just wanting some experience, on the cheap, don't even imagine such a thing.
ExtremeTech did a review of the power-line networking tech back in April 2002; they have some comparisons of its performance relative to other networking technologies; the article in question can be found here. As it stands, the powerline networking was pretty slow; even 802.11b outperformed it.
Does anyone know of any other, more recent network tech shootouts? This was the most recent I could find for powerline. Extremetech has also done some testing on Bluetooth, for anyone who is interested in how it performs.
For a detailed exposition on why the Hindenburg burned (not exploded), look here. I find it amusing that the term "Hindenburg effect" is used to describe the mindset that people have about hydrogen being dangerous.
Also, the fact that the canisters are using Metal Hydride storage is rather nice, too. Lower pressure, but heavy, and much safer. Some of the newer metal hydride materials can store the same volume of hydrogen (in low-pressure, gaseous form chemically bound to metal powder) as liquid hydrogen (without needing to mess with cryogenic materials). Ovonics recently announced a material which could store more than 7% hydrogen by weight; this means that a liter container of their metal hydride could contains about 72-73 grams of hydrogen (a liter of liquid hydrogen contains about 71 grams).
Roger Billings (currently with the International Academy of Science) who drove a hydrogen powered Cadillac in President Carter's inaugural parade (gives you an idea of WHEN), did some research on safety and metal hydride. He took some metal hydride containers, fully charged with hydrogen, to a US Army test range and had them shoot the containers with incendiary bullets. They punctured the container, and they got a "pilot light"-type flame which burned for a couple hours, but there was never a "burst" or anything remotely resembling a fireball or explosion. Safer than a tankfull of gasoline.
I ran across Plucker about six months ago. Loaded the software on my Palm M100 and my main (Linux-based) machine at home. Now, any time I run across a neat article on Slashdot, if I don't have time to read it now, I create a quick script which "plucks" it and I put it on my Palm. At that point, I can read it whenever I like.
Current contents: Contracts in Cyberspace and the VCR-faq. I was trying to fix a VCR which didn't want to play EP tapes and was always having problems with the tracking. I didn't have to bring the VCR to my computer, and I didn't have to kill any trees.
That aside, I use the daylights out of the Calendar function; medical and dental appointments for myself, my wife and my kids, when certain bills are due, when I'm expecting such-and-such check or refund, etc. It makes it a lot easier to follow-up on something when you can put it in the calendar, with a note, and have it trigger an alarm a month or more down the road.
The Memopad gets used a lot, as well. If I'm talking to the insurance agent about my vehicles, I have the VIN numbers at my fingertips, as well as the license plate numbers. I keep a file with all my passwords for the various online places where I have accounts. I also have birthdates and SSN's for my wife and kids (I'm notoriously bad at remembering dates; having an alarm go off a week prior to a birthdate or the anniversary helps).
If we're planning a roadtrip (to visit family members elsewhere), I can put together a list (in the To-Do list) of everything which needs to go, and check them off when it comes time to leave. This also comes in handy when going to the grocery store. My wife will be telling me hours beforehand, "oh, and we need this and this and this," and I can mark them all down and check them off when I pick them up.
If you have a Palm, and you haven't gotten your hands on BrainForest, you need to look at it. I tend to think in "outline fashion," and this program help tremendously. Excellent for keeping notes which are more structured and organized than you might keep with MemoPad (and you aren't as limited in the size of your file, either).
Finally, get your hands on TextPlus. It takes a little getting used to, but once you do, it definitely improves your data input speed, even in graffiti. I just wish there was a way to use it in Windows and Linux; this kind of program could actually make a pen-based system an realistic replacement for a keyboard.
I swear by my Palm. I wish I had one with more RAM (M100 = 2 MB, not expandable or upgradeable in the firmware category). I used to have a IIIx, which had more RAM and upgradability; I miss it. Also, my M100 has a problem with NOT retaining its data when the batteries come out, so I have to sync mine regularly, then restore the whole thing after I replace the batteries (rechargeable NiMH batteries). In spite of that rather large pain in the neck, I still love my Palm. I'd be a completely unorganized mess without it. I hated carrying a paper-based planner; my Palm fits in my pocket, so size really isn't an issue.
Becaue the rechargeable Lithium cells have a nominal voltage of 3.6 volts, where the AA/AAA batteries have a nominal voltage of 1.2 - 1.25 volts. Take a look here for more details.
I used to think this would be a great idea, too. I mean, since Li-Ion and Li-Polymer batteries have a higher gravimetric and volumetric energy density (wH/kg and wH/liter, respectively) than any other technology, they'd make great replacements for the AA/AAA batteries we go through. Unfortunately, not all AA/AAA consuming devices can use voltage in increments of 3.6 volts.
I remember seeing something on TV about this; they demo'ed it at Comdex.
The version they came up with (I checked the link) worked at about 2400 bps. I'd say 10 Mb/s is a substantial improvement.
Years ago, I planned my ultimate personal computer. It would be powered by my shoes, have a PAN to connect a small computing module (in contact with my skin, but discretely located) to a hand-held, tablet-style display (X protocol? something better?), and also use my cellphone to provide wireless WAN/LAM capabilities. I was thinking seriously about Bluetooth for the PAN part of the equation, but it doesn't have enough bandwidth. This might get close.
With that type of system, and some kind of integrated VoIP through the WAN/LAN link, there would never be any reason for me to actually be at my desk. In fact, it could be in my employer's best interest for me to be "on my feet."
They movie was Chain Reaction. And they found a way to make a cheap, abundant supply of hydrogen (more efficiently than traditional electrolysis methods).
Do these AP's have to have different IP addresses? If not, and you can have four wireless hops between IP addresses, you could, realistically, get some pretty decent physical distance between points were IP addresses are needed. At that rate, just about any routing protocol should be sufficient.
Slashdot Mirror
First off, the batteries are not "laptop" batteries per se. They are 18650 lithium ion cells, like these. While this ad shows them as "3.7 volts" and "1300-1400 mAh," the nominal voltage is considered 3.6 volts; let's split the difference and call them 1350 mAh. That means (3.6v x 1.350 Ah = ) 4.86 Wh / battery. A two-pack is $12.99, which mean ($12.99 / 9.72 Wh = ) $1.34 / Wh. At that rate, 50 kWh (which the article states is the car's capacity) costs about $67,000.
Their tZero gets at least 5 miles / kWh (according to other sources), which means at least 250 miles / charge (the EV1 got closer to 6 miles / kWh).
Note: many laptop battery packs (especially the third-party, aftermarket ones) are simply plastic enclosures with these replacement Li-Ion batteries in them. Hence the comment about "laptop batteries."
Still too pricey for making my own Battery Electric Vehicle (the batteries alone are more than I spent on my last three cars, combined), but we're DEFINITELY getting there. A year ago, the best deal I could find was over $11 / Wh.
If you've got RealPlayer installed on your machine, EVWorld.com has an interview with one of the car's builders, providing some more technical detail.
Oh, you mean, like some kind of tax?
I mean, the basic tenet is that if you use more gasoline, you will pay more money in taxes. And yes, under the current system, most of the money collected on those taxes is supposed to go to road repairs.
Take a look at this article. It points out that about 23% of the price of gasoline in the US is tax. In the UK, it's 76%. Any question about why petrol (gasoline) works out to about $4.37/gallon
Read an article or two about people in the UK complaining about how much it costs to drive anywhere, and think again.
The current term for that is an "air hybrid."
More details can be found here.
A couple comments on yours. You make some excellent points.
NREL typically rates "one sun" of solar flux as approx. 1 kW / square meter. Typical solar panels are, as you state, about 15% efficient. If you want an idea of how much solar flux your area gets, take a look at the maps from these pages.
The best hydrolyzers (electrolysis units) are about 60% efficient. Since a kilogram of hydrogen contains about 33 kWh of energy (if you convert at 100% efficiency), a 60% efficient hydrolyzer will need about 55 kWh to make one kilogram of hydrogen. A kilo of hydrogen has roughly the same energy content as a gallon of gasoline. Consequently, if you're going to burn hydrogen in your ICE vehicle, you'll need 55 kWh of energy to replace one gallon of gasoline. In case you're interested, that's about 14 LITERS of liquid hydrogen (density = 71 grams / liter); no other commercially available hydrogen storage can match LH2 for energy density (yeah, there are some experimental systems which have been announced, but none of the COMMERCIALLY AVAILABLE solutions are anywhere close). Go find seven 2-liter bottles of soda (or which used to contain soda), and consider hauling that volume around to replace 3.785 liters of gasoline. Any question why they're having such a hard time storing a significant amount of hydrogen?
Last, but not least, most of your mobile fuel cell stacks are only 40% efficient. That beats the ICE's by at least 50% (last time I checked, none of the ICE's are beating 25% efficiency). Let's see, 60% effeciency at the hydrolyzer, and 40% efficiency at the fuel cell equals (0.6 x 0.4 = 0.24) 24% of the electricity you fed into the hydrolyzer actually coming out of the fuel cell. Talk about wasteful.
By the time you consider how little fuel you'll be able to store, you might as well just build yourself a battery-driven electric vehicle. The GM Impact/EV II gets about 6 miles / kWh. It has a range of about 96 miles (works out to 16 kWh storage). Over 75% of the energy fed to it comes back out; beats the hell out of 24%. Also, considering the price you'd pay for the fuel cell stack (Toyota's Fuel Cell Vehicle is quoted as costing $250,000 per unit to build; they lease them, not sell them), you could probably buy NIMH or Li-Ion batteries and get some pretty impressive range.
Any argument for a fuel cell vehicle is a bigger argument for battery electric vehicles. Safety issues aside.
More details here
Fins den 'en Svenske gjente i vaerelse med flere Norske eller Danske, fin paa den peneste og du har fant den Svensker.
</language>
If there is a Swedish girl in a room with several Norwegian or Danish girls, find the prettiest and you have found the Swede.
Can't help on the Russian, though.
O.k. Where's the "gotcha" clause? I have a hard time believing that MS is actually going open-source on ANY of their IP. After all it's their "Intellectual Property," right? And if it's "property," sharing it or giving it away would reduce share value, right?
You (and various other posters on this topic) have pointed out that an FPGA provides its best advantage on hardware-accelerating certain things. Other posters have pointed out that we aren't looking at true reconfigurable computing anywhere in the next few years, as actually writing calculations to hardware is very tedious.
O.k., so let's meet somewhere in the middle. Another poster pointed out that a CPU's ISA can be thought of as an API. So, let's build an expanded API. More specifically:
There's probably a:
4. Profit!
in there somewhere.
Think about it folks; how many of us have
Just how much of libc6 would benefit from FPGA acceleration?
Well, it's big enough to get 720P (1280x720, progressive) HDTV on there and still have room for subtitles/controls, etc. at the bottom of the screen.
Find his book, "The Hydrogen World View," by Roger E. Billings (ISBN 0963163426). It's autobiographical, an amusing read, and packed with A GREAT DEAL of valuable insight.
Pegasus, gasogen, wood gas, call it whatever you want. It's all carbon monoxide, and the process is called partial oxidation.
Simple fact: heat some carbon-based material to its burning temperature in an environment which doesn't have enough oxygen. It will produce carbon monoxide (if there's enough O2, you will get CO2). Two CO molecules + one O2 molecule = CO2 + heat.
Yes, this was used heavily in France during WWII. The Germans were taking all the petrol they could get their hands on to fuel their equipment, so the civilians had to be "clever" if they wanted to still use their vehicles. Additionally, back in the '70's, some guys (Americans; saw this on Real People, if you can believe it) put a device on the back end of a small station wagon, fed it with wood scraps from lumber yards and sawmills, and drove cross country. They didn't have to pay for their fuel; most places were happy to let them take all the "waste" they wanted. Part of the wood was spent heating the oxidation chamber, part of it was going into the oxidation chamber. The resulting CO was piped under the car to the engine (a six-cylinder, IIRC). It wasn't the fastest thing on the road (you had limits to how fast you could feed it CO, so the power was kinda wimpy), but it did get out on the interstate and keep up. Considering the fact that the rest of the economy was struggling with the "energy crisis" at the time, they were trying to raise awareness to the fact that other solutions existed.
Considering the fact that Nazis were fond of using CO to kill large numbers of Jews during the Holocaust, I'd say play with this tech at your own risk.
The person to go looking for on this one is Roger Billings; he's currently with the International Academy of Science (I drive by there every morning on my way to work). He was driving hydrogen-powered ICE vehicles back in the '70's.
He ran into two problems with running an engine on Hydrogen.
He got hydrogen engines down to such an artform that he modified a Volkswagen Beetle ('72, IIRC) to run on the stuff for a college competition (he was an undergrad at the time), and the emissions coming out of the tailpipe were actually CLEANER than the air going into the intake. Basically, any Carbon Monoxide or unburned Hydrocarbons (common vehicle pollutants) which made it into the intake were finished off in the process, and the hydrogen fuel didn't produce any such emissions (water vapor).
I'd be wary about adding hydrogen to a gasoline engine to help the economy. You'd need a significant amount of hydrogen to make any real difference, and hydrogen storage these days is either:
Sorry guys, but this sounds like B.S.
Want a solution to the problem? It's called a plug-in hybrid. It runs about 20-40 miles on a charge, then kicks in an ICE to provide power after that. Result: most of the small "grocery getter" trips are done pure EV, but you can still fill up at any gas station if you go out of range.
The major auto manufacturers aren't doing this yet, but some colleges are.
Agreed. Someone mentioned a page with benchmarks of an 800 MHz, but that was an Ezra-core. The new Nehemiah core uses DDR RAM (which helps TREMENDOUSLY with the video bandwidth; the CPU doesn't use all the bandwidth, so the video uses the rest of it).
Supposedly, the newer Nehemiah-core systems are capable of playing MPEG-2 and DivX full-rate, which the older systems simply couldn't do.
This tech is much older. Take a look at this article (note: it's a .pdf file). I first read about this stuff in 1993. Texas Instruments started developing this 1983 (yes, that's two decades ago), finally abandoned it and licensed it to someone else.
Hydrazine is indeed nasty stuff. They also use it on Emergency Power Units (EPU's) on the F-16. The training they provided (for those of us who worked on the F-16) indicated that, if exposed to it, your chance of developing leukemia increases about 1,000x. Not to mention the fact that it tends to burn at about 1,600 F and is EXTREMELY caustic. We lost an F-16 (an older one; it was in the local reserve unit) while I was stationed in Utah, and the news reports REPEATEDLY told people to stay the hell away from the wreckage, due to some of the noxious chemicals on board (jet fuel isn't exactly nice stuff, but it's nothing compared to hydrazine).
Hydrazine is commonly used for such backup power systems, including on the shuttle, because it is very compact and very simple to design in. Yes, if it leaks you need people in the full chem/bio suits to clean it up. But, if properly stored and used, it is very handy stuff.
Conspiracy theories aside, folks. Get a clue.
I don't know about the eco-system part, but the phase-change aspect has me wondering.
According to the article, the external temperatures on the CPU cooler box was 24 C. According to this page, if you reduce the pressure in your water system down to 3% of 1 atm pressure (relative vacuum), that would be enough to get the water to boil. Once you got the pressure down, you probably wouldn't need any kind of pump. Simply have the water boil at the CPU cooler, then run it through a long heat-sink attached to a copper tube. The water would cool enough to condense back to a fluid, running back down the tube, and repeating the cycle.
Oh, yeah. That already exists; it's called a heat pipe. The one in the linked article is relatively small and has a fan blowing through a radiator for active cooling. How long until somebody out there "rolls their own," with a long, passive heatsink? The author of the article was looking for "quiet," and that would certainly fill the bill.
Yeah, imagine a beowulf cluster of them. When I was in college, a friend of mine was building a beowulf cluster of 386 machines. His main workstation was a 486 (note: the campus computer lab had year-old Pentium 233 machines). His entire cluster was about as fast as a machine in the campus computer lab.
Why did he do this? We wanted some experience with building, managing and programming a cluster. It cost him less than $200 to build a 10 machine cluster, and he could honestly say he had some experience with it. When he wanted to get something done quickly, however, he walked to the campus computer lab and used one of their machines. Unless you're just wanting some experience, on the cheap, don't even imagine such a thing.
ExtremeTech did a review of the power-line networking tech back in April 2002; they have some comparisons of its performance relative to other networking technologies; the article in question can be found here. As it stands, the powerline networking was pretty slow; even 802.11b outperformed it.
Does anyone know of any other, more recent network tech shootouts? This was the most recent I could find for powerline. Extremetech has also done some testing on Bluetooth, for anyone who is interested in how it performs.
For a detailed exposition on why the Hindenburg burned (not exploded), look here. I find it amusing that the term "Hindenburg effect" is used to describe the mindset that people have about hydrogen being dangerous.
Also, the fact that the canisters are using Metal Hydride storage is rather nice, too. Lower pressure, but heavy, and much safer. Some of the newer metal hydride materials can store the same volume of hydrogen (in low-pressure, gaseous form chemically bound to metal powder) as liquid hydrogen (without needing to mess with cryogenic materials). Ovonics recently announced a material which could store more than 7% hydrogen by weight; this means that a liter container of their metal hydride could contains about 72-73 grams of hydrogen (a liter of liquid hydrogen contains about 71 grams).
Roger Billings (currently with the International Academy of Science) who drove a hydrogen powered Cadillac in President Carter's inaugural parade (gives you an idea of WHEN), did some research on safety and metal hydride. He took some metal hydride containers, fully charged with hydrogen, to a US Army test range and had them shoot the containers with incendiary bullets. They punctured the container, and they got a "pilot light"-type flame which burned for a couple hours, but there was never a "burst" or anything remotely resembling a fireball or explosion. Safer than a tankfull of gasoline.
I ran across Plucker about six months ago. Loaded the software on my Palm M100 and my main (Linux-based) machine at home. Now, any time I run across a neat article on Slashdot, if I don't have time to read it now, I create a quick script which "plucks" it and I put it on my Palm. At that point, I can read it whenever I like.
Current contents: Contracts in Cyberspace and the VCR-faq. I was trying to fix a VCR which didn't want to play EP tapes and was always having problems with the tracking. I didn't have to bring the VCR to my computer, and I didn't have to kill any trees.
That aside, I use the daylights out of the Calendar function; medical and dental appointments for myself, my wife and my kids, when certain bills are due, when I'm expecting such-and-such check or refund, etc. It makes it a lot easier to follow-up on something when you can put it in the calendar, with a note, and have it trigger an alarm a month or more down the road.
The Memopad gets used a lot, as well. If I'm talking to the insurance agent about my vehicles, I have the VIN numbers at my fingertips, as well as the license plate numbers. I keep a file with all my passwords for the various online places where I have accounts. I also have birthdates and SSN's for my wife and kids (I'm notoriously bad at remembering dates; having an alarm go off a week prior to a birthdate or the anniversary helps).
If we're planning a roadtrip (to visit family members elsewhere), I can put together a list (in the To-Do list) of everything which needs to go, and check them off when it comes time to leave. This also comes in handy when going to the grocery store. My wife will be telling me hours beforehand, "oh, and we need this and this and this," and I can mark them all down and check them off when I pick them up.
If you have a Palm, and you haven't gotten your hands on BrainForest, you need to look at it. I tend to think in "outline fashion," and this program help tremendously. Excellent for keeping notes which are more structured and organized than you might keep with MemoPad (and you aren't as limited in the size of your file, either).
Finally, get your hands on TextPlus. It takes a little getting used to, but once you do, it definitely improves your data input speed, even in graffiti. I just wish there was a way to use it in Windows and Linux; this kind of program could actually make a pen-based system an realistic replacement for a keyboard.
I swear by my Palm. I wish I had one with more RAM (M100 = 2 MB, not expandable or upgradeable in the firmware category). I used to have a IIIx, which had more RAM and upgradability; I miss it. Also, my M100 has a problem with NOT retaining its data when the batteries come out, so I have to sync mine regularly, then restore the whole thing after I replace the batteries (rechargeable NiMH batteries). In spite of that rather large pain in the neck, I still love my Palm. I'd be a completely unorganized mess without it. I hated carrying a paper-based planner; my Palm fits in my pocket, so size really isn't an issue.
Becaue the rechargeable Lithium cells have a nominal voltage of 3.6 volts, where the AA/AAA batteries have a nominal voltage of 1.2 - 1.25 volts. Take a look here for more details.
I used to think this would be a great idea, too. I mean, since Li-Ion and Li-Polymer batteries have a higher gravimetric and volumetric energy density (wH/kg and wH/liter, respectively) than any other technology, they'd make great replacements for the AA/AAA batteries we go through. Unfortunately, not all AA/AAA consuming devices can use voltage in increments of 3.6 volts.
I remember seeing something on TV about this; they demo'ed it at Comdex.
The version they came up with (I checked the link) worked at about 2400 bps. I'd say 10 Mb/s is a substantial improvement.
Years ago, I planned my ultimate personal computer. It would be powered by my shoes, have a PAN to connect a small computing module (in contact with my skin, but discretely located) to a hand-held, tablet-style display (X protocol? something better?), and also use my cellphone to provide wireless WAN/LAM capabilities. I was thinking seriously about Bluetooth for the PAN part of the equation, but it doesn't have enough bandwidth. This might get close.
With that type of system, and some kind of integrated VoIP through the WAN/LAN link, there would never be any reason for me to actually be at my desk. In fact, it could be in my employer's best interest for me to be "on my feet."
They movie was Chain Reaction. And they found a way to make a cheap, abundant supply of hydrogen (more efficiently than traditional electrolysis methods).
Quick question.
Do these AP's have to have different IP addresses? If not, and you can have four wireless hops between IP addresses, you could, realistically, get some pretty decent physical distance between points were IP addresses are needed. At that rate, just about any routing protocol should be sufficient.