Tuesday, November 24, 2009
New Theory of Gravity Decouples Space and Time
"Petr Horava, a physicist at the University of California in Berkeley, has a new theory about gravity and spacetime. At high energies, it actually snips any ties between space and time, yet at low energies devolves to equivalence with the theory of General Relativity, which binds them together. The theory is gaining popularity with physicists because it fits some observations better than Einstein's or Newton's solutions. It better predicts the movement of the planets (in an idealized case) and has a potential to create the illusion of dark matter. Another physicist calculated that under Horava Gravity, our universe would experience not a Big Bang but a Big Bounce — and the new theory reproduces the ripples from such an event in a way that matches measurements of the cosmic microwave background."
Wednesday, November 11, 2009
Star Trek-like Replicator? Electron Beam Device Makes Metal Parts, One Layer At A Time
A group of engineers working on a novel manufacturing technique at NASA's Langley Research Center in Hampton, Va., have come up with a new twist on the popular old saying about dreaming and doing: "If you can slice it, we can build it."
That's because layers mean everything to the environmentally-friendly construction process called Electron Beam Freeform Fabrication, or EBF3, and its operation sounds like something straight out of science fiction.
"You start with a drawing of the part you want to build, you push a button, and out comes the part," said Karen Taminger, the technology lead for the Virginia-based research project that is part of NASA's Fundamental Aeronautics Program.
She admits that, on the surface, EBF3 reminds many people of a Star Trek replicator in which, for example, Captain Picard announces out loud, "Tea, Earl Grey, hot." Then there is a brief hum, a flash of light and the stimulating drink appears from a nook in the wall.
In reality, EBF3 works in a vacuum chamber, where an electron beam is focused on a constantly feeding source of metal, which is melted and then applied as called for by a drawing -- one layer at a time -- on top of a rotating surface until the part is complete.
While the options for using EBF3 are more limited than what science fiction allows, the potential for the process is no less out of this world, with promising relevance in aviation, spaceflight -- even the medical community, Taminger said.
Commercial applications for EBF3 are already known and its potential already tested, Taminger said, noting it's possible that, within a few years, some aircraft will be flying with large structural parts made by this process.
To make EBF3 work there are two key requirements: A detailed three-dimensional drawing of the object to be created must be available, and the material the object is to be made from must be compatible for use with an electron beam.
First, the drawing is needed to break up the object into layers, with each cross-section used to guide the electron beam and source of metal in reproducing the object, building it up layer by layer.
"If you take a slice through a typical truss, you can see a couple of dots in each cross-section that move as you go from layer to layer," Taminger said. "When complete, you see those moving dots actually allowed you to build a diagonal brace into the truss."
Second, the material must be compatible with the electron beam so that it can be heated by the stream of energy and briefly turned into liquid form, making aluminum an ideal material to be used, along with other metals.
In fact, the EBF3 can handle two different sources of metal -- also called feed stock -- at the same time, either by mixing them together into a unique alloy or embedding one material inside another.
The potential use for the latter could include embedding a strand of fiber optic glass inside an aluminum part, enabling the placement of sensors in areas that were impossible before, Taminger said.
While the EBF3 equipment tested on the ground is fairly large and heavy, a smaller version was created and successfully test flown on a NASA jet that is used to provide researchers with brief periods of weightlessness. The next step is to fly a demonstration of the hardware on the International Space Station, Taminger said.
Future lunar base crews could use EBF3 to manufacture spare parts as needed, rather than rely on a supply of parts launched from Earth. Astronauts might be able to mine feed stock from the lunar soil, or even recycle used landing craft stages by melting them.
But the immediate and greatest potential for the process is in the aviation industry where major structural segments of an airliner, or casings for a jet engine, could be manufactured for about $1,000 per pound less than conventional means, Taminger said.
Environmental savings also are made possible by deploying EBF3, she added.
Normally an aircraft builder might start with a 6,000-pound block of titanium and machine it down to a 300-pound part, leaving 5,700 pounds of material that needs to be recycled and using several thousand gallons of cutting fluid used in the process..
"With EBF3 you can build up the same part using only 350 pounds of titanium and machine away just 50 pounds to get the part into its final configuration," Taminger said. "And the EBF3 process uses much less electricity to create the same part."
While initial parts for the aviation industry will be simple shapes, replacing parts already designed, future parts designed from scratch with the EBF3 process in mind could lead to improvements in jet engine efficiency, fuel burn rate and component lifetime.
"There's a lot of power in being able to build up your part layer by layer because you can get internal cavities and complexities that are not possible with machining from a solid block of material," Taminger said.
That's because layers mean everything to the environmentally-friendly construction process called Electron Beam Freeform Fabrication, or EBF3, and its operation sounds like something straight out of science fiction.
"You start with a drawing of the part you want to build, you push a button, and out comes the part," said Karen Taminger, the technology lead for the Virginia-based research project that is part of NASA's Fundamental Aeronautics Program.
She admits that, on the surface, EBF3 reminds many people of a Star Trek replicator in which, for example, Captain Picard announces out loud, "Tea, Earl Grey, hot." Then there is a brief hum, a flash of light and the stimulating drink appears from a nook in the wall.
In reality, EBF3 works in a vacuum chamber, where an electron beam is focused on a constantly feeding source of metal, which is melted and then applied as called for by a drawing -- one layer at a time -- on top of a rotating surface until the part is complete.
While the options for using EBF3 are more limited than what science fiction allows, the potential for the process is no less out of this world, with promising relevance in aviation, spaceflight -- even the medical community, Taminger said.
Commercial applications for EBF3 are already known and its potential already tested, Taminger said, noting it's possible that, within a few years, some aircraft will be flying with large structural parts made by this process.
To make EBF3 work there are two key requirements: A detailed three-dimensional drawing of the object to be created must be available, and the material the object is to be made from must be compatible for use with an electron beam.
First, the drawing is needed to break up the object into layers, with each cross-section used to guide the electron beam and source of metal in reproducing the object, building it up layer by layer.
"If you take a slice through a typical truss, you can see a couple of dots in each cross-section that move as you go from layer to layer," Taminger said. "When complete, you see those moving dots actually allowed you to build a diagonal brace into the truss."
Second, the material must be compatible with the electron beam so that it can be heated by the stream of energy and briefly turned into liquid form, making aluminum an ideal material to be used, along with other metals.
In fact, the EBF3 can handle two different sources of metal -- also called feed stock -- at the same time, either by mixing them together into a unique alloy or embedding one material inside another.
The potential use for the latter could include embedding a strand of fiber optic glass inside an aluminum part, enabling the placement of sensors in areas that were impossible before, Taminger said.
While the EBF3 equipment tested on the ground is fairly large and heavy, a smaller version was created and successfully test flown on a NASA jet that is used to provide researchers with brief periods of weightlessness. The next step is to fly a demonstration of the hardware on the International Space Station, Taminger said.
Future lunar base crews could use EBF3 to manufacture spare parts as needed, rather than rely on a supply of parts launched from Earth. Astronauts might be able to mine feed stock from the lunar soil, or even recycle used landing craft stages by melting them.
But the immediate and greatest potential for the process is in the aviation industry where major structural segments of an airliner, or casings for a jet engine, could be manufactured for about $1,000 per pound less than conventional means, Taminger said.
Environmental savings also are made possible by deploying EBF3, she added.
Normally an aircraft builder might start with a 6,000-pound block of titanium and machine it down to a 300-pound part, leaving 5,700 pounds of material that needs to be recycled and using several thousand gallons of cutting fluid used in the process..
"With EBF3 you can build up the same part using only 350 pounds of titanium and machine away just 50 pounds to get the part into its final configuration," Taminger said. "And the EBF3 process uses much less electricity to create the same part."
While initial parts for the aviation industry will be simple shapes, replacing parts already designed, future parts designed from scratch with the EBF3 process in mind could lead to improvements in jet engine efficiency, fuel burn rate and component lifetime.
"There's a lot of power in being able to build up your part layer by layer because you can get internal cavities and complexities that are not possible with machining from a solid block of material," Taminger said.
Tuesday, November 3, 2009
Portege 2000 Laptop, SD Card OS (Experiment)
Been working on a Portege 2000 Laptop. Started by pulling the hard drive completely out. This reduces noise of course as well as battery life. The problem was, what are you left with. A laptop that will only boot to the bios. To install an operating system is already difficult enough on a Portege 2000 as the laptop comes with no cd-rom drive or any other internal drive. It is equipped with an SD card slot though. Like what camera's and cellphones use to store ringtones etc. The goal was to make an SD card my hard drive. This would allow awesome battery life as well as NO MOVING PARTS. No spinning hard drive, no noise, and hopefully a fast read write speed to the flash memory.
I had previously booted pendrive linux from a network via PXE and TTFTP on the same laptop, so I knew it could be done. The pendrive linux was booted via PXE over a network as the portege has no bios settings for usb boot or even supports for external cd-rom drive or floppy drive. The problem was pendrive linux nor dsl linux could properly install due to no hard drive storage. I could trick them into running though. Those distros need something to copy to.
After playing with a 4 gig CF card and 1.8" to CF adapter, I could get the bios to recognize the CF, but not able to install to it. I partitioned to FAT,FAT32,and then NTFS with no avail. Still was not able to install to the disk. I gave up on that idea and moved to the forgotten SD card. I had a 128mb a 512mb and a 2 gig stick to play with.
The hard part was making the SD bootable. After multiple tries, I managed to install EEEUbuntu on the 2 gig SD card, setting aside the extra 1200mb or so for the persistant drive. This allows you to save stuff to the drive. Got it bootable via DOSOSOAR and an executable .bat file. The Portege does not support booting via usb so I had to PXE boot from a network and load a plbt boot file or (GRUB) to allow various boot options. I plugged in the network cable, started the ttftp server, set the portege to boot via usb and she fired up. Load time on the Ubuntu was a little slow. The SD card speeds are not up to say a 7200 rpm hard drive but I have no moving parts. To top that off I can remove the SD Card, put it in my pocket, take it to any other computer and boot my stuff on it. This make the portege a paperweight. Without a hard drive, and SD card in my pocket, the laptop is useless to a theif. I have all my info in my pocket, they have a useless piece of metal.
After configuring Ubuntu speeds started increasing. I still have caching on the SD set up so I dont expect the actual life of the card to last long. You would really need support for UDMA to increase life on the read write side. The principle is there though. I am working on making the display through xrandr scripts and manipulating the /etc/x11/xorg.conf file to set the display to 1024x768_60.00 for some reason the ubuntu default must be 800x600_59
I have started by adding the new mode with this string #!xrandr --newmode "1024x768_60.00" etc. etc. etc. <--(diff modes)but have been unsuccessful due to the permissions. Ubuntu has something screwy with the root logon and allowing permission to edit some files. The #!chmod 755 /etc/x11/xorg.conf still will not allow me to get access.
Overall, the whole process has been fun. Other than the display problems, ubuntu is a great distro. It supported my audio, mouse, etc. and even recognized the usb wifi adapter right off. Still running tests on the battery life and possibly making the SD a little faster. I have read multiple forums stating that a computer cannot boot nor run from SD. It is the same principle as booting from usb, just with a twist. SUPERGEEK 5000 OUT!
I had previously booted pendrive linux from a network via PXE and TTFTP on the same laptop, so I knew it could be done. The pendrive linux was booted via PXE over a network as the portege has no bios settings for usb boot or even supports for external cd-rom drive or floppy drive. The problem was pendrive linux nor dsl linux could properly install due to no hard drive storage. I could trick them into running though. Those distros need something to copy to.
After playing with a 4 gig CF card and 1.8" to CF adapter, I could get the bios to recognize the CF, but not able to install to it. I partitioned to FAT,FAT32,and then NTFS with no avail. Still was not able to install to the disk. I gave up on that idea and moved to the forgotten SD card. I had a 128mb a 512mb and a 2 gig stick to play with.
The hard part was making the SD bootable. After multiple tries, I managed to install EEEUbuntu on the 2 gig SD card, setting aside the extra 1200mb or so for the persistant drive. This allows you to save stuff to the drive. Got it bootable via DOSOSOAR and an executable .bat file. The Portege does not support booting via usb so I had to PXE boot from a network and load a plbt boot file or (GRUB) to allow various boot options. I plugged in the network cable, started the ttftp server, set the portege to boot via usb and she fired up. Load time on the Ubuntu was a little slow. The SD card speeds are not up to say a 7200 rpm hard drive but I have no moving parts. To top that off I can remove the SD Card, put it in my pocket, take it to any other computer and boot my stuff on it. This make the portege a paperweight. Without a hard drive, and SD card in my pocket, the laptop is useless to a theif. I have all my info in my pocket, they have a useless piece of metal.
After configuring Ubuntu speeds started increasing. I still have caching on the SD set up so I dont expect the actual life of the card to last long. You would really need support for UDMA to increase life on the read write side. The principle is there though. I am working on making the display through xrandr scripts and manipulating the /etc/x11/xorg.conf file to set the display to 1024x768_60.00 for some reason the ubuntu default must be 800x600_59
I have started by adding the new mode with this string #!xrandr --newmode "1024x768_60.00" etc. etc. etc. <--(diff modes)but have been unsuccessful due to the permissions. Ubuntu has something screwy with the root logon and allowing permission to edit some files. The #!chmod 755 /etc/x11/xorg.conf still will not allow me to get access.
Overall, the whole process has been fun. Other than the display problems, ubuntu is a great distro. It supported my audio, mouse, etc. and even recognized the usb wifi adapter right off. Still running tests on the battery life and possibly making the SD a little faster. I have read multiple forums stating that a computer cannot boot nor run from SD. It is the same principle as booting from usb, just with a twist. SUPERGEEK 5000 OUT!
Time-travel doesn't imbue quantum computers with superpowers
After spending the past two months on sabbatical, I've returned to a deluge of science that I had missed out on. Almost three years ago, a company called D-Wave made waves by announcing that it was about to unveil one of the first-ever functioning adiabatic quantum computers, a device it heralded as being capable of solving NP problems in P time—a claim that doesn't really hold up to scrutiny.
This piqued the interest of many about the actual applications and science behind quantum computers, and we at Nobel Intent dove in and tried to shine some light on the discussion. Even with a fully functional, scalable quantum computer, nobody's large integers are in danger of being factored in polynomial time—it has been shown that integer factorization, via Shor's algorithm, is solvable in bounded error quantum polynomial (BQP) time. So, quantum mechanicists and quantum computer theorists started looking for ways to improve upon the performance of quantum computers and arrived at a question only a theorist could come up with. What if the quantum computer was capable of traveling through time?
A paper that was published in the October 21st edition of Physical Review Letters (PRL) examines this very question. The writers attempt to see if a quantum computer that exists on a closed timelike circuit (CTC)—a timeline that travels to the past before looping back onto itself—experiences an increase in its computational ability. A previous paper in PRL this year suggested that it was possible for a CTC-assisted quantum computer to map a set of pure states into an orthogonal set of states, an impossibility in standard quantum mechanics. This would imply that a CTC would allow a quantum computer to distinguish two identical quantum states—the philosophical or physical meaning and implication of this being entirely unclear.
Aware that something must have been wrong—either with the procedure or assumptions of the previous work—the authors of this paper from the IBM Watson Research Laboratory and the Quantum Computing Department at the University of Waterloo revisit the problem with a highly critical and pedantic eye. Their new analysis shows that "CTCs do not improve state discrimination," contradicting what was previously reported. The problem now becomes how to reconcile these results. The answer lies in the fact that 2 + 2 doesn't always add up to 4.
In a linear system, the properties of a mixture are simply equal to the (weighted) sum of the properties of the individual components. In fact, quantum mechanics is mathematically founded on the idea of a linear set of equations and linear independence of solutions. It turns out that a CTC does not represent a linear operation, rather a nonlinear one where the outcome is not merely the sum of the components. Computationally, this means that a quantum computer traveling through a loop in time would only see a computational benefit with a specific subset of inputs, not the more general case of every possible input as has been previous postulated. For that general case to be possible, the operation of the computer looping in time would need to be linear.
The fact that quantum mechanics could have some nonlinearities has been soundly rebutted in past literature, but this is a new proposal which may shed new light onto the subject. It would also change the way we need to think about quantum problems: not only would the particles in a given system need to be considered, but every particle in the universe, even those that do not participate, would need to be taken into account. It would also mean that a complex problem of two or more systems would need to take into account the entire history of the universe that each subsystem exists in from the begining of time to the present to accurately calculate a result.
While this paper has, in my opinion, an excellently mind-bending postulate, it is not something that one needs to lose sleep over. It does highlight (again, in my opinion) why theory is so much more elegant than experimental work. Where else can entire papers and fields of study be carried out where no realistic quantum computer has ever been developed, and no method of time travel has ever been shown to be physically possible (note that is also has not been shown to be physically impossible), yet can work out what would happen if they had a time traveling quantum computer? Awesome.
This piqued the interest of many about the actual applications and science behind quantum computers, and we at Nobel Intent dove in and tried to shine some light on the discussion. Even with a fully functional, scalable quantum computer, nobody's large integers are in danger of being factored in polynomial time—it has been shown that integer factorization, via Shor's algorithm, is solvable in bounded error quantum polynomial (BQP) time. So, quantum mechanicists and quantum computer theorists started looking for ways to improve upon the performance of quantum computers and arrived at a question only a theorist could come up with. What if the quantum computer was capable of traveling through time?
A paper that was published in the October 21st edition of Physical Review Letters (PRL) examines this very question. The writers attempt to see if a quantum computer that exists on a closed timelike circuit (CTC)—a timeline that travels to the past before looping back onto itself—experiences an increase in its computational ability. A previous paper in PRL this year suggested that it was possible for a CTC-assisted quantum computer to map a set of pure states into an orthogonal set of states, an impossibility in standard quantum mechanics. This would imply that a CTC would allow a quantum computer to distinguish two identical quantum states—the philosophical or physical meaning and implication of this being entirely unclear.
Aware that something must have been wrong—either with the procedure or assumptions of the previous work—the authors of this paper from the IBM Watson Research Laboratory and the Quantum Computing Department at the University of Waterloo revisit the problem with a highly critical and pedantic eye. Their new analysis shows that "CTCs do not improve state discrimination," contradicting what was previously reported. The problem now becomes how to reconcile these results. The answer lies in the fact that 2 + 2 doesn't always add up to 4.
In a linear system, the properties of a mixture are simply equal to the (weighted) sum of the properties of the individual components. In fact, quantum mechanics is mathematically founded on the idea of a linear set of equations and linear independence of solutions. It turns out that a CTC does not represent a linear operation, rather a nonlinear one where the outcome is not merely the sum of the components. Computationally, this means that a quantum computer traveling through a loop in time would only see a computational benefit with a specific subset of inputs, not the more general case of every possible input as has been previous postulated. For that general case to be possible, the operation of the computer looping in time would need to be linear.
The fact that quantum mechanics could have some nonlinearities has been soundly rebutted in past literature, but this is a new proposal which may shed new light onto the subject. It would also change the way we need to think about quantum problems: not only would the particles in a given system need to be considered, but every particle in the universe, even those that do not participate, would need to be taken into account. It would also mean that a complex problem of two or more systems would need to take into account the entire history of the universe that each subsystem exists in from the begining of time to the present to accurately calculate a result.
While this paper has, in my opinion, an excellently mind-bending postulate, it is not something that one needs to lose sleep over. It does highlight (again, in my opinion) why theory is so much more elegant than experimental work. Where else can entire papers and fields of study be carried out where no realistic quantum computer has ever been developed, and no method of time travel has ever been shown to be physically possible (note that is also has not been shown to be physically impossible), yet can work out what would happen if they had a time traveling quantum computer? Awesome.
Subscribe to:
Posts (Atom)