Not remotely close to the first time someone stitched multiple TEM images together. This reminds me of the (in)famous 1994 article "A Mathematical Model for the Determination of Total Area Under Glucose Tolerance and Other Metabolic Curves", in which a medical doctor discovered numerical integration.
While the LHC is a big deal, cooling things to 1.9 K isn't. People have been doing it for decades. In fact, there are more than a handful of physicists who routinely work with things held at tens of mK.
Moore's Law has nothing to do with mechanical components. For that matter, strictly speaking, all Moore's Law was ever about was the number of components in an IC. Not speed, not cost, not even size. And definitely not the size of MEMS!
I think the example of London is a strong counter-example. Charge people enough and give them relatively viable transportation alternatives and they DO stop driving into at least one urban center.
The document says subjects were exposed to 1.4 W/kg - that's a huge amount of power compared to a transmitting cell phone - at least an order of magnitude more, if not two orders.
Just because you don't succesfully identify someone doesn't mean you get a false positive. The best system on the market successfully identifies people over 99.9% of the time and has never had a false positive in several billion searches.
Uh, most glass (certainly that used for eye-glasses) is pretty damn transmissive in the near-IR (700-900 nm), where almost all (if not all) iris systems operate. As another example, if glass was strongly absorbing in the IR, optical fiber communication wouldn't work particularly well. Fiber optics typically operate at 1310, 1550, or 850 nm.
MIT Lincoln Labs (244 Wood St, Lexington, MA)
GE Global Research Center (1 Research Circle, Niskayuna, NY)
Knolls Atomic Power Lab (next door to GE, formerly operated by GE)
I work for a biometrics company. There was one product on the market a number of years ago that used a laser. However, one can image the retina for biometric purposes without a laser. In addition, the concept has NEVER been to scar the eye - you simply look at the blood vessels, which are clearly visible. This is no different than with finger and palm vein systems which use infrared to look at blood vessels under the skin. Retinal imaging has a number of advantages from a security point of view, among them being nearly spoof-proof in practice.
One piece of relevant information: in my case, I only ever see my collaborators once or twice a year (they're overseas). Face-to-face meetings can be poorly used and overused, but being able to meet in person on a somewhat regular basis can be extremely useful.
You are right that whether or not telecommuting is productive or not depends upon circumstances. For my work, which often involves very detailed technical discussion, it can be very helpful to actually have the person you're talking to in front of you, be able to draw on a whiteboard, play with the modeling softeare he's using and so on. This is not to say taht there aren't situations where a competent, independent person can get the job done perfectly well from a remote location.
I don't see anyone asking the question: "what effect does telecommuting have on productivity?" I work in the R&D arm of a major multinational corporation and the projects I work on are highly collaborative. I can often accomplish more in 15-30 minutes of face-to-face conversation with a colleague than in an hour or more over the phone or video conference, even with fancy collaboration tools like Lotus Sametime and Microsoft NetMeeting.
I'm not so sure this is the biggest problem. This is the first time I've heard of edge-localized modes being a huge problem (granted, I am not a plasma physicist). Most times I've seen people raising practical concerns about large tokamaks for energy production, the "biggest" problem cited has been neutron bombardment of the reactor walls. Energetic neutrons have the nasty habit of making the vessel walls radioactive and - worse - making them brittle and prone to mechanical failure.
The whole idea behind "sonofusion" or "bubble fusion" is that the pressure wave created by an imploding bubble creates a small region of very, very high temperature, which, combined with the pressure of implosion is sufficient to initiate fusion. The first paper was subject to much controversy, but the results were ultimately reproduced by others and objections regarding the instrumentation addressed.
It hasn't happened yet, but nuclear fusion would be quite green. The small quantities of activated material produced by neutron bombardment would be a small price to pay. Never mind the potential for aneutronic reactors which would produce no radioactive byproducts.
Two comments:
1. Apple had done serious GUI work BEFORE coming to PARC (cf. the article on Larry Tesler in the latest IEEE Spectrum magazine). Also, by the time of the PARC visit, folks like Jef Raskin had already been at Apple for some time.
2. Apple toured PARC in exchange for PAYING Xerox for the privelege.
This is "absolutely and totally" wrong. First of all, a major portion of the Cell architecture is a 64-bit PowerPC core, flanked by assorted other DSP units. Second, IBM is the major architect of the Cell, which was not designed ex nihilo.
See this Ars review: http://arstechnica.com/articles/paedia/cpu/cell-1. ars
There have been dozens of fusion reactors, if not more over the last few decades, some of them in operation now. The NYT has posted an expanded article that corrects the gaffe. Some names you can google for more info:
TFTR
NSTX
JET
C-Mod
DIII-D
It's not likely to make things that much cheaper. One of the major costs for current (193nm) lithography is the need for an excimer laser source. Liquid immersion isn't going to change that. Also, this technique will still require clean rooms similar to those in use now. It's a clever technique to prevent optical lithography from running out of gas just yet, but it's merely delaying the inevitable. Which, of course, is what keeps the industry profitable - so this is not insignificant. It's not mind blowing either, though.
The article is unclear. All they're talking about is using liquid immersion optics. So rather than shrinking the wavelength, which can no longer be done without switching to reflective optics, they're increasing the numerical apeture by imaging through a liquid. This idea has been around for a while and is nothing new. It still does not address the exponentially increasing costs of optical lithogrpaphy tools - in fact liquid immersion optics will only complicate matters further and cotinue to drive costs up.
Not to be trite, but these things take time.
MRAM, OLEDs, and similar promising technologies often respresent a fundamental shift in technology. This means new manufacturing processes, extensive testing, etc. The industry has been doing the transistor thing for a long time now and they know what they're doing. They're naturally reluctant to change the way they do things regardless of whether there's a clear technological advantage or not. What it takes is a shot at making boatloads beyond their wildest dreams OR the solution to a immediate, critical problem. (Part of) the reason the MOSFET really took off is because bipolar transistors were starting to consume ridiculous amounts of power (sort of where we are now, actually). Switching to CMOS almost immediately bought the industry 20 more years of wiggle room.
Meanwhile, transistor-based RAM isn't killing anybody, so it will be a while before alternative technologies take off.
A 1.2 GHz G5 only dissipates 19W, which puts it in the ballpark of feasibility in a laptop/notebook. There are clearly marketing concerns, though - they already have a 1.33 GHz G4's and they're selling and they probably want to wait for a notebook-ready 2 GHz G5 (or somethhing similar) before putting out a G5 Powerbook.
Slashdot Mirror
Not remotely close to the first time someone stitched multiple TEM images together. This reminds me of the (in)famous 1994 article "A Mathematical Model for the Determination of Total Area Under Glucose Tolerance and Other Metabolic Curves", in which a medical doctor discovered numerical integration.
While the LHC is a big deal, cooling things to 1.9 K isn't. People have been doing it for decades. In fact, there are more than a handful of physicists who routinely work with things held at tens of mK.
Moore's Law has nothing to do with mechanical components. For that matter, strictly speaking, all Moore's Law was ever about was the number of components in an IC. Not speed, not cost, not even size. And definitely not the size of MEMS!
I think the example of London is a strong counter-example. Charge people enough and give them relatively viable transportation alternatives and they DO stop driving into at least one urban center.
The document says subjects were exposed to 1.4 W/kg - that's a huge amount of power compared to a transmitting cell phone - at least an order of magnitude more, if not two orders.
Blue LEDs are made with Galium Nitride.
Just because you don't succesfully identify someone doesn't mean you get a false positive. The best system on the market successfully identifies people over 99.9% of the time and has never had a false positive in several billion searches.
Uh, most glass (certainly that used for eye-glasses) is pretty damn transmissive in the near-IR (700-900 nm), where almost all (if not all) iris systems operate. As another example, if glass was strongly absorbing in the IR, optical fiber communication wouldn't work particularly well. Fiber optics typically operate at 1310, 1550, or 850 nm.
MIT Lincoln Labs (244 Wood St, Lexington, MA) GE Global Research Center (1 Research Circle, Niskayuna, NY) Knolls Atomic Power Lab (next door to GE, formerly operated by GE)
What you're talking about is simply one retinal imaging system. See Retica Systems for a company that has a non-contact, non-invasive system.
I work for a biometrics company. There was one product on the market a number of years ago that used a laser. However, one can image the retina for biometric purposes without a laser. In addition, the concept has NEVER been to scar the eye - you simply look at the blood vessels, which are clearly visible. This is no different than with finger and palm vein systems which use infrared to look at blood vessels under the skin. Retinal imaging has a number of advantages from a security point of view, among them being nearly spoof-proof in practice.
One piece of relevant information: in my case, I only ever see my collaborators once or twice a year (they're overseas). Face-to-face meetings can be poorly used and overused, but being able to meet in person on a somewhat regular basis can be extremely useful.
You are right that whether or not telecommuting is productive or not depends upon circumstances. For my work, which often involves very detailed technical discussion, it can be very helpful to actually have the person you're talking to in front of you, be able to draw on a whiteboard, play with the modeling softeare he's using and so on. This is not to say taht there aren't situations where a competent, independent person can get the job done perfectly well from a remote location.
I don't see anyone asking the question: "what effect does telecommuting have on productivity?" I work in the R&D arm of a major multinational corporation and the projects I work on are highly collaborative. I can often accomplish more in 15-30 minutes of face-to-face conversation with a colleague than in an hour or more over the phone or video conference, even with fancy collaboration tools like Lotus Sametime and Microsoft NetMeeting.
Vrms = sqrt(3kT/m). For oxygen, at T=300K, I get less than 700 m/s. according to Google, escape velocity for the moon is about 2400 m/s.
I'm not so sure this is the biggest problem. This is the first time I've heard of edge-localized modes being a huge problem (granted, I am not a plasma physicist). Most times I've seen people raising practical concerns about large tokamaks for energy production, the "biggest" problem cited has been neutron bombardment of the reactor walls. Energetic neutrons have the nasty habit of making the vessel walls radioactive and - worse - making them brittle and prone to mechanical failure.
The whole idea behind "sonofusion" or "bubble fusion" is that the pressure wave created by an imploding bubble creates a small region of very, very high temperature, which, combined with the pressure of implosion is sufficient to initiate fusion. The first paper was subject to much controversy, but the results were ultimately reproduced by others and objections regarding the instrumentation addressed.
It hasn't happened yet, but nuclear fusion would be quite green. The small quantities of activated material produced by neutron bombardment would be a small price to pay. Never mind the potential for aneutronic reactors which would produce no radioactive byproducts.
Two comments: 1. Apple had done serious GUI work BEFORE coming to PARC (cf. the article on Larry Tesler in the latest IEEE Spectrum magazine). Also, by the time of the PARC visit, folks like Jef Raskin had already been at Apple for some time. 2. Apple toured PARC in exchange for PAYING Xerox for the privelege.
This is "absolutely and totally" wrong. First of all, a major portion of the Cell architecture is a 64-bit PowerPC core, flanked by assorted other DSP units. Second, IBM is the major architect of the Cell, which was not designed ex nihilo. See this Ars review: http://arstechnica.com/articles/paedia/cpu/cell-1. ars
There have been dozens of fusion reactors, if not more over the last few decades, some of them in operation now. The NYT has posted an expanded article that corrects the gaffe. Some names you can google for more info: TFTR NSTX JET C-Mod DIII-D
It's not likely to make things that much cheaper. One of the major costs for current (193nm) lithography is the need for an excimer laser source. Liquid immersion isn't going to change that. Also, this technique will still require clean rooms similar to those in use now. It's a clever technique to prevent optical lithography from running out of gas just yet, but it's merely delaying the inevitable. Which, of course, is what keeps the industry profitable - so this is not insignificant. It's not mind blowing either, though.
The article is unclear. All they're talking about is using liquid immersion optics. So rather than shrinking the wavelength, which can no longer be done without switching to reflective optics, they're increasing the numerical apeture by imaging through a liquid. This idea has been around for a while and is nothing new. It still does not address the exponentially increasing costs of optical lithogrpaphy tools - in fact liquid immersion optics will only complicate matters further and cotinue to drive costs up.
Not to be trite, but these things take time. MRAM, OLEDs, and similar promising technologies often respresent a fundamental shift in technology. This means new manufacturing processes, extensive testing, etc. The industry has been doing the transistor thing for a long time now and they know what they're doing. They're naturally reluctant to change the way they do things regardless of whether there's a clear technological advantage or not. What it takes is a shot at making boatloads beyond their wildest dreams OR the solution to a immediate, critical problem. (Part of) the reason the MOSFET really took off is because bipolar transistors were starting to consume ridiculous amounts of power (sort of where we are now, actually). Switching to CMOS almost immediately bought the industry 20 more years of wiggle room. Meanwhile, transistor-based RAM isn't killing anybody, so it will be a while before alternative technologies take off.
A 1.2 GHz G5 only dissipates 19W, which puts it in the ballpark of feasibility in a laptop/notebook. There are clearly marketing concerns, though - they already have a 1.33 GHz G4's and they're selling and they probably want to wait for a notebook-ready 2 GHz G5 (or somethhing similar) before putting out a G5 Powerbook.