If there's one thing I learned from taking classes about nuclear fusion, it's this: generating net fusion power is difficult. This goes for magnetic as well as inertial confinement schemes. So I caution you against being too enthusiastic about any particular initiative. The history of fusion research is a pattern of "oh this is a great idea, we'll have it in 10 years", followed by "uhh there are all sorts of unexpected issues". Progress is slow and painstaking, and TANSTAAFL is the rule. Don't be surprised when previously inexpensive concepts turn out to require significant investment to achieve net power production.
But I agree that we should fund research across the board.
The main issues for ITER are in the realm of plasma physics and materials science.
Currently, ITER is predicted to work based on arguments from "wind-tunnel scaling": make it bigger and certain figures of merit improve. This scaling is based on magnetohydrodynamics (MHD), but until the darn thing is built there's no way to know for sure whether the predictions are correct. Furthermore, there are "advanced modes" that aren't fully understood from a theoretical standpoint.
The "first wall", the inside wall of the vacuum vessel, is the thorniest problem in terms of materials. It has to both withstand an intense neutron flux, and avoid sloughing off bits that contaminate the plasma. Similar requirements are necessary for the "divertors", which sit in contact with the plasma and kind of hold it in place.
It's important to note that the ITER project is not just the reactor; the associated International Fusion Materials Irradiation Facility is key to resolving these tough questions. We've come through the Space Age with some pretty neat materials, but compared to what's required for fusion, they look like child's play.
Assuming there's decent energy gain, the challenge for ICF becomes the rate. For laser ICF, you need a rate on the order of 1 Hz (generously). It's going to take a while before that becomes possible. And when the rate goes up, the neutrons will have to be dealt with.
Thinking about it a bit more, in comparison, the ideal temperature for DT fusion is 15 keV = 174,000,000 K. I don't know what the pp fusion cross section vs temperature looks like, but since it's not in the tables of the NRL Plasma Formulary it's probably not worth pursuing.
Even at temperatures in the sun's core, 15,000,000 Kelvins or 27,000,000 Fahrenheit, the average lifetime of a proton against pp fusion is about 8,000,000,000 years.
In the mean time, there are a number of other very promising approaches which continue to be neglected, and these could be funded at a small fraction of the cost. What is very frustrating is that most of these have been around for a long time, and some were even cancelled so that our futile pursuit of Tokamaks could continue.
But seriously, with the hedging language in the statement you've quoted, there's nothing controversial. Note the "up to" and "aiming ultimately". (Plus "prolonged" in this line of business means a few minutes.) Fusion scientists are cautious people, having made rosy predictions in the past that never came to fruition. And when you're cautious, it's hard to convince lawmakers to hand over the money.
On the other hand, ITER as a concept has been around since the '80s. If they had just gone ahead with it back then, we would have learned a lot by now. Same goes for the cancellation of the SSC.
You don't need general relativity for most situations. That includes most of the solar system.
You may be thinking of GPS's reliance on relativity, which is only with regards to timekeeping: both the relative velocity of the satellites and their distance from the earth's gravity well result in time dilation.
Standard orbital mechanics works fine for satellites, as long as they are not in a very strong gravitational field. The classic example is the perihelion of Mercury, where results based on Newtonian theory are erroneous, due to Mercury's proximity to the sun.
The problem is, with high-speed rail we'd be buying our track technology and rolling stock from Japan or Germany, since it makes no sense to start from scratch. So much for national pride.
In my experience the bags aren't always big enough to hold hard drives. TI mails me samples via Digikey, and AD sends me little plastic boxes with foam.
Unless you know something I don't, you can't make a silicon wire smaller than the width of a single atom, so there is definitely a physical limit that we aren't that far away from.
Sure, physical limitations exist. Can't argue with nature. But people are looking at building upward, in 3D rather than 2D. That will keep the transistor count going up.
... at 3.2 GHz, the electromagnetic waves inside the chip can only propagate about 4.7 centimeters per cycle.
I've seen this argument before but it's not going to be relevant anytime soon. The capacitance of interconnect is a bigger limitation than the speed of light. And once the speed of light becomes an issue, pipelining will keep it at bay for a while.
You've probably noticed that we haven't had any really major jumps in the clock speeds of consumer processors since about 2002. Intel originally thought they'd be able to scale the Pentium 4 Netburst architecture to about 10GHz, bu they ran into a frequency ceiling at about 4GHz.
Netburst was a serious misstep for Intel, so I'm not sure why you'd want to use it as an example. Clock speed has not scaled much because the market has changed its focus. Performance per clock has increased, and there is a focus on performance per watt. Clock-for-clock, the Athlon64/Opteron wiped the floor with Netburst, and set the standard for performance/watt. Core2 similarly took another step forward, while also increasing the clock speed a bit with the Penryn shrink. Core i7 does more per clock than the Core2, not to mention per watt.
Netburst also illustrates how seemingly fatal impediments to scaling are defeated through engineering. Prescott at 90nm had a big problem with leakage, and at smaller process nodes the leakage was expected to be an even greater issue. Fast forward a few years to Penryn, with its high-k metal gate tech. Problem solved, and Moore's Law marches on.
In short, unless there is a major materials breakthrough, or materials change, I would expect Moore's law to hold for the next five years or so, but not much longer after that. We're rapidly approaching the physical limits.
Moore's law involves transistor count, not clock speed. Note the graph in the WP article.
But I agree that the infernal P4 got the industry to rethink clock speed as the be-all and end-all of microprocessing. Leakage at 90nm and below was a big problem, too.
People have been saying for years that we're about to reach the end of the line in terms of Moore's law. So far they've all been proven wrong, and scaling continues unabated.
Dumping processors in a box is "easy", but multicore programming is not easy. The software tools are not there yet. Not to mention, you need deep pockets to roll your own multicore IC and build up the requisite software ecosystem. Just look at how much trouble Sony had with Cell. Everybody is watching to see if Intel will succeed with Larrabee.
Now Oracle may have good reason to be interested in Sun's Niagara. Database applicances, perhaps.
And where does Apple come into this, exactly? PA Semi's focus is on a totally different market segment.
I've not tried the Woodford Reserve myself, but it did win a Double Gold and "Best Bourbon" at the SF Spirits competition, so it better be a good sipping Bourbon.;) denttford says that it has a smoky characteristic, so I agree that it probably doesn't have the right flavor profile for mixed drinks.
In the center-of-mass frame, the reaction products are distributed isotropically. (Probably; neutrinos are not my area of expertise.) However, the center-of-mass frame is moving relativistically in the direction the proton beam is going. Hence in the lab frame, the resulting neutrino beam will favor a certain direction.
Slashdot Mirror
Now that I look at it, it's a pretty impressive list:
http://www.matroska.org/technical/specs/tagging/index.html
Somehow I've never felt the need to embed the metadata in the video file itself. I usually make an .nfo to go along with it.
I've been using mkv for years, but never thought about its lack of support for metadata. That's an interesting point.
If there's one thing I learned from taking classes about nuclear fusion, it's this: generating net fusion power is difficult. This goes for magnetic as well as inertial confinement schemes. So I caution you against being too enthusiastic about any particular initiative. The history of fusion research is a pattern of "oh this is a great idea, we'll have it in 10 years", followed by "uhh there are all sorts of unexpected issues". Progress is slow and painstaking, and TANSTAAFL is the rule. Don't be surprised when previously inexpensive concepts turn out to require significant investment to achieve net power production.
But I agree that we should fund research across the board.
The main issues for ITER are in the realm of plasma physics and materials science.
Currently, ITER is predicted to work based on arguments from "wind-tunnel scaling": make it bigger and certain figures of merit improve. This scaling is based on magnetohydrodynamics (MHD), but until the darn thing is built there's no way to know for sure whether the predictions are correct. Furthermore, there are "advanced modes" that aren't fully understood from a theoretical standpoint.
The "first wall", the inside wall of the vacuum vessel, is the thorniest problem in terms of materials. It has to both withstand an intense neutron flux, and avoid sloughing off bits that contaminate the plasma. Similar requirements are necessary for the "divertors", which sit in contact with the plasma and kind of hold it in place.
It's important to note that the ITER project is not just the reactor; the associated International Fusion Materials Irradiation Facility is key to resolving these tough questions. We've come through the Space Age with some pretty neat materials, but compared to what's required for fusion, they look like child's play.
Assuming there's decent energy gain, the challenge for ICF becomes the rate. For laser ICF, you need a rate on the order of 1 Hz (generously). It's going to take a while before that becomes possible. And when the rate goes up, the neutrons will have to be dealt with.
Thinking about it a bit more, in comparison, the ideal temperature for DT fusion is 15 keV = 174,000,000 K. I don't know what the pp fusion cross section vs temperature looks like, but since it's not in the tables of the NRL Plasma Formulary it's probably not worth pursuing.
Pretty darned difficult:
http://www.tim-thompson.com/fusion.html
In the mean time, there are a number of other very promising approaches which continue to be neglected, and these could be funded at a small fraction of the cost. What is very frustrating is that most of these have been around for a long time, and some were even cancelled so that our futile pursuit of Tokamaks could continue.
Such as?
In the words of wikipedia, citation please?
http://dx.doi.org/10.1016/j.jnucmat.2004.04.004
But seriously, with the hedging language in the statement you've quoted, there's nothing controversial. Note the "up to" and "aiming ultimately". (Plus "prolonged" in this line of business means a few minutes.) Fusion scientists are cautious people, having made rosy predictions in the past that never came to fruition. And when you're cautious, it's hard to convince lawmakers to hand over the money.
On the other hand, ITER as a concept has been around since the '80s. If they had just gone ahead with it back then, we would have learned a lot by now. Same goes for the cancellation of the SSC.
I leave my Gmail session open in Chrome, and use Firefox for most everything else.
It sounds difficult to run a 1-km grid simulation on today's hardware, let alone 15 years ago.
You don't need general relativity for most situations. That includes most of the solar system.
You may be thinking of GPS's reliance on relativity, which is only with regards to timekeeping: both the relative velocity of the satellites and their distance from the earth's gravity well result in time dilation.
Standard orbital mechanics works fine for satellites, as long as they are not in a very strong gravitational field. The classic example is the perihelion of Mercury, where results based on Newtonian theory are erroneous, due to Mercury's proximity to the sun.
The problem is, with high-speed rail we'd be buying our track technology and rolling stock from Japan or Germany, since it makes no sense to start from scratch. So much for national pride.
In my experience the bags aren't always big enough to hold hard drives. TI mails me samples via Digikey, and AD sends me little plastic boxes with foam.
When I was looking at entry-level EE job listings a few years ago, most of them specified a MS. It made me, with just a BS, a bit sad.
Unless you know something I don't, you can't make a silicon wire smaller than the width of a single atom, so there is definitely a physical limit that we aren't that far away from.
Sure, physical limitations exist. Can't argue with nature. But people are looking at building upward, in 3D rather than 2D. That will keep the transistor count going up.
... at 3.2 GHz, the electromagnetic waves inside the chip can only propagate about 4.7 centimeters per cycle.
I've seen this argument before but it's not going to be relevant anytime soon. The capacitance of interconnect is a bigger limitation than the speed of light. And once the speed of light becomes an issue, pipelining will keep it at bay for a while.
You've probably noticed that we haven't had any really major jumps in the clock speeds of consumer processors since about 2002. Intel originally thought they'd be able to scale the Pentium 4 Netburst architecture to about 10GHz, bu they ran into a frequency ceiling at about 4GHz.
Netburst was a serious misstep for Intel, so I'm not sure why you'd want to use it as an example. Clock speed has not scaled much because the market has changed its focus. Performance per clock has increased, and there is a focus on performance per watt. Clock-for-clock, the Athlon64/Opteron wiped the floor with Netburst, and set the standard for performance/watt. Core2 similarly took another step forward, while also increasing the clock speed a bit with the Penryn shrink. Core i7 does more per clock than the Core2, not to mention per watt.
Netburst also illustrates how seemingly fatal impediments to scaling are defeated through engineering. Prescott at 90nm had a big problem with leakage, and at smaller process nodes the leakage was expected to be an even greater issue. Fast forward a few years to Penryn, with its high-k metal gate tech. Problem solved, and Moore's Law marches on.
In short, unless there is a major materials breakthrough, or materials change, I would expect Moore's law to hold for the next five years or so, but not much longer after that. We're rapidly approaching the physical limits.
They'll think of something.
Moore's law involves transistor count, not clock speed. Note the graph in the WP article.
But I agree that the infernal P4 got the industry to rethink clock speed as the be-all and end-all of microprocessing. Leakage at 90nm and below was a big problem, too.
I think Niagara may fit the bill.
People have been saying for years that we're about to reach the end of the line in terms of Moore's law. So far they've all been proven wrong, and scaling continues unabated.
Dumping processors in a box is "easy", but multicore programming is not easy. The software tools are not there yet. Not to mention, you need deep pockets to roll your own multicore IC and build up the requisite software ecosystem. Just look at how much trouble Sony had with Cell. Everybody is watching to see if Intel will succeed with Larrabee.
Now Oracle may have good reason to be interested in Sun's Niagara. Database applicances, perhaps.
And where does Apple come into this, exactly? PA Semi's focus is on a totally different market segment.
FOX is the only channel that provides both the "big government is the only solution" and the "government is the problem" viewpoints.
Don't you realize they are promoting a false dichotomy? Fox shows the world in black and white.
You can set Firefox to open PDFs externally instead of using the plugin. Options -> Applications.
That's what I did until last week, when I switched to Sumatra.
I've not tried the Woodford Reserve myself, but it did win a Double Gold and "Best Bourbon" at the SF Spirits competition, so it better be a good sipping Bourbon. ;) denttford says that it has a smoky characteristic, so I agree that it probably doesn't have the right flavor profile for mixed drinks.
I suggest reading these articles, but as always, there's a lot more to explore:
http://www.nytimes.com/2007/11/28/dining/28bour.html
http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2003/01/02/WI144547.DTL
http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2008/12/19/WIBC14LRGR.DTL
I tend to drink more Rye than anything else, though.
http://www.nytimes.com/2006/11/29/dining/29wine.html
http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2007/03/16/WIRYE.DTL
In the center-of-mass frame, the reaction products are distributed isotropically. (Probably; neutrinos are not my area of expertise.) However, the center-of-mass frame is moving relativistically in the direction the proton beam is going. Hence in the lab frame, the resulting neutrino beam will favor a certain direction.
So does the host interface part reside in the Lattice FPGA, in 10K LUTs?
The PIC requires some infrastructure, though: compiler/assembler, programmer. The 555 requires no external programming.
That said, it's amazing what you can do with a dirt-cheap microcontroller these days.