I agree that there is no simple answer, but I am uneasy with your "experiment is larger than the data" concept. Today we think of the Michelson-Morley and Millikan experiments as canonical and definitive investigations in Physics. But we do not often remember that each was preceded by a string of less-successful experiments, and followed by confirmations. It the accumulation of a body of data that leads to the gradual acceptance of a physical concept.
Let's stop right there. There are no general lessons to be had from the LHC. It's an exception, not the rule.
First: 99.9% of scientists are not working at LHC, or any other billion dollar, world-unique facility. They are working in ordinary labs, with ordinary equipment that's identical or similar to equipment in hundreds of other labs around the world.
I admit that I jumped on the LHC as an extreme example. But even in an "ordinary" lab these days, you'll find some specialized and complex equipment. This is true for the cutting edge of any field.
Second: Primary data, actual measurement results, are already kept, as a rule.
As oneiros27 notes, this is not guaranteed, either by design or circumstance.
Third: The vast majority of experiments are never ever reproduced to begin with. You're lucky enough to get cited, really. Most papers don't even get cited apart from by those who wrote them.
Not sure what kind of point you're trying to make here.
Fourth: Very little science is done by re-interpreting existing results. That only applies to the unique cases where the actual experiment can't be reproduced easily.
It's not necessarily a matter of re-interpreting existing results. You may be adding an old dataset to a new dataset, and finding new results in the combined set, or finding a glimmer of something new in an old dataset. Even for "small" experiments, having somebody else's raw dataset can make your life a lot easier.
What happens when five years later it's discovered that a flawed assumption was used in the analysis? Are we going to build another LHC any time soon, to verify the result?
Truth is, you'd still have to rebuild the LHC then, because you didn't test your 'corrected' assumption against the actual machine to show that your 'corrected' results are valid. Until the actual experiment is re-done it'll remain an unanswered question.
No, I am talking strictly about analysis. For example, the use of neural networks in particle/track finding has recently met greater acceptance in in High Energy Physics. But what happens if, a few years down the road, evidence turns up that neural networks are fundamentally flawed? If you have kept the data, you can re-run the analysis with different methods. If you have thrown out the data, it's time to build a new LHC.
Granted, High Energy Physics, with its requirements for large datasets in order to find extremely rare processes, is perhaps the only branch of science to require so much data. In HEP, we want to keep as much as possible, but there are realistic limits. In other fields, since there are no difficulties, why not keep everything?
You seem to be using "results" in a wider sense than "published papers". Yes, nobody is going to throw out papers. But the raw data from instruments? It is not clear whether those will be kept.
You say that the analysis and interpretations can be thrown out, but those portions are precisely what go into published papers. And for small-scale science, it makes little sense to throw away anything at all.
Let's say the LHC publishes its analysis, and then throws away the data. What happens when five years later it's discovered that a flawed assumption was used in the analysis? Are we going to build another LHC any time soon, to verify the result?
For a billion-dollar experiment like the LHC, that dataset is the prize. The dataset is the whole reason the LHC was built. Physicists will be combing the data for rare events and odd occurrences, many years down the road.
Corporations can come up with enough capital, say, to build a power plant. That's just a few billions dollars and some land.
But face it, some things are just so expensive and large-scale that no private companies are going to bet tens of billions of dollars on them, for a payoff decades down the road. A private company will build a toll road, but not an entire Interstate Highway System.
The speedup is not always of that magnitude. If what you do is embarrassingly parallel, then yes, it will be a good candidate for an FPGA. However these days, processing cycles are cheap and getting cheaper, so it does not always make sense to do things in hardware.
The thing about Snow Crash is that the Metasphere is a completely different place, and your RL body doesn't have to go anywhere while you're linked up. TFA is postulating a move toward something like the ubiquitous glasses in Dennou Coil, where a digital overlay is projected over the real world.
From what I remember, a fusion-fission hybrid combines the difficultly of maintaining a fusion reaction with the volatility of high-level waste. Not really a good idea.
If we really want to burn off high-level waste, it makes more sense to me to go the direction of Rubbia's Energy Amplifier. We're within an order of magnitude of the beam power required.
The abstract mentions "confirming the presence of true Bose-Einstein macroscopic coherence (BEC) of cavity exciton polaritons." Can somebody elucidate?
I recall that the original (animated) Cowboy Bebop movie was only average -- it was basically a long episode -- so this live-action movie is going to be tough to do right. Them picking Keanu shows that they're already on the wrong track.
I agree that the source presents some difficulties. Cowboy Bebop is convoluted over the entire series (as it should be, IMHO) but episode by episode it's simple and... episodic. In a long series you can drop little hints here and there before pulling out the backstory, but in a movie there's only so much you can do without losing focus overall.
My expectation is that the movie will be crap, but perhaps they'll leave the soundtrack alone. It would be nice to have cleaner versions of "Tank!" and "Rush", and a more solid rendition of "Adieu". However "The Real Folk Blues" is in Japanese, so I expect Hollywood to replace it with something screwy.
I'm in despair. Hollywood and Keanu Reeves have left me in despair.
You are assuming an isotropic emitter, where field strength falls off as 1/r^2. That behavior is invalid for other antennas; for example a dipole's field strength falls off as 1/r (in the far-field approximation). The paper is complicated by the fact that the radiation patterns of the antennas used in this paper are directional and different. The "conventional" chip used a folded dipole with a "boresight radiation pattern", and the "proposed" chip used a custom design with a front-to-back ratio of 10dB.
Table 1 has the numbers: Module Type / Power Consumption / Gain / Range
Standalone TX chip / 3.3 mW / -34 dBi / 1 m
TX chip in conventional LTCC package / 38 mW / -1 dBi / 75 m
TX chip in proposed LTCC package / 3.3 mW / -2.3 dBi / 24 m
Let's do some reckless hand-wavy extrapolation. The difference in power is 38/3.3 = 11.5 = 10.6 dB; if we assume perfect scaling of the new package to 38mW, we'd expect 10.6-2.3=8.3 dBi. This is an improvement of 9.3 dB over the conventional method -- it's almost 10 times as efficient.
This analysis ignores, among other things, the relative directionalities of the antennas. I wonder why they didn't choose a more directional antenna for the "conventional" chip, or used the same sort of antenna in order to do a level comparison.
The other point of comparison is between the "standalone" chip and the "proposed" chip. A 32 dB improvement with no power increase is nothing to sneeze at!
Slashdot Mirror
OK :) I'll add that I read Snow Crash for the first time just last year and was blown away by how fresh it seemed, years after it was first published.
I agree that there is no simple answer, but I am uneasy with your "experiment is larger than the data" concept. Today we think of the Michelson-Morley and Millikan experiments as canonical and definitive investigations in Physics. But we do not often remember that each was preceded by a string of less-successful experiments, and followed by confirmations. It the accumulation of a body of data that leads to the gradual acceptance of a physical concept.
See chart:
http://en.wikipedia.org/wiki/Michelson-Morley_experiment#The_most_famous_failed_experiment
Let's stop right there. There are no general lessons to be had from the LHC. It's an exception, not the rule.
First: 99.9% of scientists are not working at LHC, or any other billion dollar, world-unique facility.
They are working in ordinary labs, with ordinary equipment that's identical or similar to equipment in hundreds of other labs around the world.
I admit that I jumped on the LHC as an extreme example. But even in an "ordinary" lab these days, you'll find some specialized and complex equipment. This is true for the cutting edge of any field.
Second: Primary data, actual measurement results, are already kept, as a rule.
As oneiros27 notes, this is not guaranteed, either by design or circumstance.
Third: The vast majority of experiments are never ever reproduced to begin with. You're lucky enough to get cited, really. Most papers don't even get cited apart from by those who wrote them.
Not sure what kind of point you're trying to make here.
Fourth: Very little science is done by re-interpreting existing results. That only applies to the unique cases where the actual experiment can't be reproduced easily.
It's not necessarily a matter of re-interpreting existing results. You may be adding an old dataset to a new dataset, and finding new results in the combined set, or finding a glimmer of something new in an old dataset. Even for "small" experiments, having somebody else's raw dataset can make your life a lot easier.
Truth is, you'd still have to rebuild the LHC then, because you didn't test your 'corrected' assumption against the actual machine to show that your 'corrected' results are valid. Until the actual experiment is re-done it'll remain an unanswered question.
No, I am talking strictly about analysis. For example, the use of neural networks in particle/track finding has recently met greater acceptance in in High Energy Physics. But what happens if, a few years down the road, evidence turns up that neural networks are fundamentally flawed? If you have kept the data, you can re-run the analysis with different methods. If you have thrown out the data, it's time to build a new LHC.
Granted, High Energy Physics, with its requirements for large datasets in order to find extremely rare processes, is perhaps the only branch of science to require so much data. In HEP, we want to keep as much as possible, but there are realistic limits. In other fields, since there are no difficulties, why not keep everything?
You seem to be using "results" in a wider sense than "published papers". Yes, nobody is going to throw out papers. But the raw data from instruments? It is not clear whether those will be kept.
You say that the analysis and interpretations can be thrown out, but those portions are precisely what go into published papers. And for small-scale science, it makes little sense to throw away anything at all.
Let's say the LHC publishes its analysis, and then throws away the data. What happens when five years later it's discovered that a flawed assumption was used in the analysis? Are we going to build another LHC any time soon, to verify the result?
For a billion-dollar experiment like the LHC, that dataset is the prize. The dataset is the whole reason the LHC was built. Physicists will be combing the data for rare events and odd occurrences, many years down the road.
Corporations can come up with enough capital, say, to build a power plant. That's just a few billions dollars and some land.
But face it, some things are just so expensive and large-scale that no private companies are going to bet tens of billions of dollars on them, for a payoff decades down the road. A private company will build a toll road, but not an entire Interstate Highway System.
So, Intel processors use microcode patches. Does that make them "not really hardware"?
The speedup is not always of that magnitude. If what you do is embarrassingly parallel, then yes, it will be a good candidate for an FPGA. However these days, processing cycles are cheap and getting cheaper, so it does not always make sense to do things in hardware.
Perhaps it was a temporary friendship, minus an accent mark.
No one will dispute that Halo brought a lot to the console arena. To a PC gamer, though, it was underwhelming.
The thing about Snow Crash is that the Metasphere is a completely different place, and your RL body doesn't have to go anywhere while you're linked up. TFA is postulating a move toward something like the ubiquitous glasses in Dennou Coil, where a digital overlay is projected over the real world.
That's because Jeeves knows that proper martinis are made with gin.
Androids have parents?
You know, the book was pretty superficial.
No, I think that's a stereotype. For instance, look at the Amazon reviews for Richard Nisbett's The Geography of Thought, a book that purports to show different ways of thinking between East and West.
http://www.amazon.com/Geography-Thought-Asians-Westerners-Differently/dp/0743255356/
Frankly, I'm worried that I'm going to be upstaged by all the smart people in China who are working a lot harder than I am.
From what I remember, a fusion-fission hybrid combines the difficultly of maintaining a fusion reaction with the volatility of high-level waste. Not really a good idea.
If we really want to burn off high-level waste, it makes more sense to me to go the direction of Rubbia's Energy Amplifier. We're within an order of magnitude of the beam power required.
The abstract mentions "confirming the presence of true Bose-Einstein macroscopic coherence (BEC) of cavity exciton polaritons." Can somebody elucidate?
I work at LBL (as a guest, not an employee) and Steven Chu is very well-liked around here. He does have a rather disturbing laugh though.
As a Bay Area resident, let me say:
Help! Help! I'm being oppressed!
But seriously, it's not that bad.
I recall that the original (animated) Cowboy Bebop movie was only average -- it was basically a long episode -- so this live-action movie is going to be tough to do right. Them picking Keanu shows that they're already on the wrong track.
I agree that the source presents some difficulties. Cowboy Bebop is convoluted over the entire series (as it should be, IMHO) but episode by episode it's simple and ... episodic. In a long series you can drop little hints here and there before pulling out the backstory, but in a movie there's only so much you can do without losing focus overall.
My expectation is that the movie will be crap, but perhaps they'll leave the soundtrack alone. It would be nice to have cleaner versions of "Tank!" and "Rush", and a more solid rendition of "Adieu". However "The Real Folk Blues" is in Japanese, so I expect Hollywood to replace it with something screwy.
I'm in despair. Hollywood and Keanu Reeves have left me in despair.
I've been reading James Gleick's Chaos, which is a pop-sci sort of book. I'll definitely check out Strogatz when I've got the time.
I had a KT7A back in the day, and now I'm running an IP-35 Pro. Good boards, and it's sad to see the company go.
It would be nice if they could release BIOS documentation, but I guess that's highly unlikely.
I'd bump the fusion budget from $300M to $500M
Let's just hope that the US fulfills its ITER contribution this time around.
You are assuming an isotropic emitter, where field strength falls off as 1/r^2. That behavior is invalid for other antennas; for example a dipole's field strength falls off as 1/r (in the far-field approximation). The paper is complicated by the fact that the radiation patterns of the antennas used in this paper are directional and different. The "conventional" chip used a folded dipole with a "boresight radiation pattern", and the "proposed" chip used a custom design with a front-to-back ratio of 10dB.
Table 1 has the numbers:
Module Type / Power Consumption / Gain / Range
Standalone
TX chip / 3.3 mW / -34 dBi / 1 m
TX chip in
conventional
LTCC package / 38 mW / -1 dBi / 75 m
TX chip in
proposed LTCC
package / 3.3 mW / -2.3 dBi / 24 m
Let's do some reckless hand-wavy extrapolation. The difference in power is 38/3.3 = 11.5 = 10.6 dB; if we assume perfect scaling of the new package to 38mW, we'd expect 10.6-2.3=8.3 dBi. This is an improvement of 9.3 dB over the conventional method -- it's almost 10 times as efficient.
This analysis ignores, among other things, the relative directionalities of the antennas. I wonder why they didn't choose a more directional antenna for the "conventional" chip, or used the same sort of antenna in order to do a level comparison.
The other point of comparison is between the "standalone" chip and the "proposed" chip. A 32 dB improvement with no power increase is nothing to sneeze at!
Greenland melt seems to be picking up speed