According to that list, the worst quake they've suffered in hundreds of years is a 7.2. Sure, the Haiti quake caused widespread damage at only a 7.0, but that's because they're dirt poor and living in shacks. With proper preparation, a 7.2 really isn't bad at all. Seismic isolation foundations have been in use for over two millennia. Building such structures on the containment building and reactor core would make it easily capable of handling such a quake.
I'm surprised no one has yet commented to this effect, but why would you want to use this patent? As I read it, the patent is for a very simple feedback control system for positioning of heliostats (mirrors). You put a camera on the collector, pointed at the mirror, and the camera controls the alignment of the mirror to center the point of highest intensity (the sun). Seems simple enough.
The first problem, this only works for a single mirror. That means you would need one of these light intensity sensors for each individual mirror, of which there may number thousands. Each of these is going to be on the collector, potentially blocking a significant amount of light from those mirrors. Now you could put a single camera on a rotating boom, allowing it to move around and individually manage each mirror in sequence, but that's still an overly complicated system.
You know the layout of your plant, or at least you should. Why not just use a single camera, tracking the sun across the sky, and use that combined with a bit of geometry to determine the optimum placement of each mirror to follow it. The other system has the advantage of being able to track the source of highest intensity, but surely any other source of light will be inconsequential compared to the sun. The next closest object (the moon) at its brightest might only provide a few kW of power to a several hundred MW plant.
But wait! There's more! The sun is a celestial object, and celestial objects are nothing if not predictable. Why bother with cameras at all? A nominal amount of CPU power would be able to predict the sun's track across the sky with micro-arcsecond accuracy. There's absolutely no need for any sort of feedback system at all, besides the position sensors built into the servo motors themselves. This just seems like Google had some image processing expertise, and decided to throw science at the wall to see what would stick.
This isn't a voice/video chat specification. This is a framework for direct browser to browser communication, using a central server as a negotiator. Google Talk (jabber) would then be built on top of this framework.
The issue is that it requires flash, and is only implemented in flash. If you want to use it, you have to write your application in flash. Similarly, Skype only works with the Skype client, and if you want to use their P2P voice network, you have to use their software. The difference here is that its an open specification. Anyone is free to implement it however they choose, and are not constrained to one company's support of a closed binary application.
a microwave could be up to 1300 watts and the full force of one will cook you quite quickly).
Oh come on. Even in open air, with no shielding, the power density of a 1.3kW magnetron falls below that of normal sunlight at somewhere under two feet.
Wait. I was under the impression that the GM plants were engineered to be incapable of pollination, to require farmers to purchase new seeds every crop. Clearly if they are pollinating to other fields, the plant in question is no longer the patented technology in question
Actually they make gas turbine engines for (expensive) RC aircraft that put out between 50-100 lbs of thrust, or you could build your own out of a high end automotive turbocharger. The issue is that you're right back to the big ducted fan, trading a largish two-cycle reciprocating engine for a smaller turbine. The OP was talking about a jetpack of similar size to a hikers pack, which would mean you would be using those RC turbines directly.
Actually ducted fans are jets. The term jet simply refers to an entrained flow inside a surrounding medium. Ducted fans produce jets with a rotor in a cowling. Pumpjets produce jets with an impeller in a cowling. Rivers produce jets as they flow into the ocean. Your garden hose produces a jet when you open the valve.
A turbojet is one specific type of jet produced by the exhaust of a gas turbine engine. A turbofan is another type of jet produced by shaft output of a gas turbine engine powering a ducted fan. There is nothing inherently tying a gas turbine engine to a jet.
Two stroke engines don't run anything close to that efficiency, and the more highly stressed they are, the more unburned oil you mix into the fuel for lubrication.
Strap a couple of those miniature jet engines (you'll probably need four for your weight, their weight, and a decent amount of fuel) to your back side, and if you don't burn your legs off, you'll at least start a brush fire. As simple as you make the concept out to be, the reason no one has ever bothered to do that is because no one has figured out a way to resolve the problem of exhaust temperature.
Meanwhile a large ducted fan means significantly lesser power requirement, and similarly, significantly less waste heat you have to deal with.
There are three factors you have to consider when building something like this: thrust to weight, thrust to power, and jet temperature. Your thrust to weight needs to be above unity, for obvious reasons. The most efficient way to achieve that thrust is a very large fan, at a very low flow velocity. As you try to shrink the unit, your fan becomes smaller, you need higher and higher flow velocity to achieve the same thrust. As you trade size for power, you go from reciprocating ducted fan, to high bypass turbofan, to low bypass turbofan, to turbojet. Each step means reduced efficiency and significantly higher exit temperature.
What you ask is certainly do-able with a turbojet, however your exhaust would start brush fires, melt asphalt, burn off your legs, and depending on how long you stayed near the ground, suffocate you and eventually itself. Now before you say 'but but but... current jetpacks are that small', that's because current jetpacks are all cold gas rockets. They have a tank of hydrogen peroxide and a catalyst. The catalyst causes rapid disassociation, and the resultant jet is a mixture of water vapor and oxygen. There is no combustion to generate heat, and as a result, they are ridiculously inefficient, which is why you get operation times of under a minute.
Alpha and beta radiation is radiation for as long as it is actually radiating. As soon as it impacts a surface and sticks, it becomes helium and electrons.
Radiation is short lived, and not a contaminate you can simply remove. Isotopes undergoing decay to produce said radiation can be removed.
Certainly with the PS3, the Cell was not an off-the-shelf design, but was an entirely new architecture, requiring entirely new compiler research. Now with the PS4, the architecture is stable, the compiler research has been done. Process migration, and duplication for multi-core, carries its own difficulties, but those are relatively simple compared to the initial development. It shouldn't take nearly the amount of time scaling the chip for 4-8 times the performance of the original Cell, if by no other means than by adding more parallelism. Sony funded the Cell expecting it would run their whole multimedia lineup for the next 20 years. I don't expect them to abandon all those costs and switch to a new processor.
While there are certainly production limits with XDR, graphics memory is plentiful and cheap. GDDR5 would work just fine as a replacement high speed store on a next generation Cell, of course there would be a fair bit of development spent replacing the existing memory controller.
Off-the-shelf CPUs/GPUs, or custom ASICs using 3rd-party licensed CPU/GPU designs (instead of designing one from scratch)
The last two generations of consoles have used off-the-shelf GPUs. At most, they have the manufacturer bring in some new tricks bound for their next generation as a boost. They do not design one from scratch, and there is no reason for them to consider doing so in the future.
The Cell has already been designed. It is now an off-the-shelf processor. There is no need to design a new architecture from scratch. Take the original Cell, increase size up the ring bus, double the SPEs, and then put 4-6 of those modules on a single die with cache and a shared memory controller. By the time the PS4 is released, IBM will be on a 22nm production process, and the new chip will actually be smaller, cheaper, and use less power than the original.
Off-the-shelf DDR(1/2/3/4/5/whatever) SDRAM (instead of using something from Rambus)
I only partially agree with this one. With the kind of floating point performance the Cell was offering, standard DDR2 at the time simply didn't cut it. DDR3 didn't support those kinds of speeds until over two years after the release of the PS3. As it stands, the Cell is pretty heavily tied into Rambus interfaces, and I don't know if that could be worked around. IMHO, they should use ~1GB of XDR2 or some form of GDDR at 100GB/s+, for the same reasons as before. Then tack on another 4-8GB of cheap DDR3. It's a travesty that the current consoles are stuck with so little memory. Optical mediums are slow. Rotating disks are slow. Use a shit load of memory and use it as a data prefetch to get rid of any form of loading time. Dump anything you need in there for the next couple minutes, and then use the high speed GDDR/XDR as a low speed cache.
Blu-ray, instead of a new kind of optical disk design (or, even eliminate the physical medium altogether in favor of online purchases)
The new optical format is already designed and pushed into mass market. The most they might want to consider would be switching to a 4-layer design, which really doesn't cost a whole lot.
Sony has done the same thing with the PS3. There have been about a dozen different hardware revisions since the original two release units, with multiple die shrinks on the silicon. Production costs are probably under $200, and power consumption is about a third what it originally was.
The problem is that the keyboard is designed for typing, not command input, and the existing chords are not well laid out. For instance, you use ctrl+tab to cycle through your tabs. Now you want to close one. With the mouse gesture, it's a click and a flick of the wrist to close it. It takes moving your whole arm to contort your hand into the shape needed to hit ctrl+f4. Now a Nostromo or similar programmable keypad would work great in this regard, but then you're no longer talking about controls that are standard to an OS or UI paradigm.
Personally, I don't understand why the number pad didn't shift to the left side of the keyboard for exactly this reason 20 years ago when the mouse started becoming popular. Unless you enjoy one-handed slow typing, you've got to move your hand over all that dead space to get it onto the character section. Conversely, if you actually do want to use the numpad for control or number entry, you either have to use your right hand, leaving your left hand is sitting idle and useless, or you need to have a foot and a half extra space to shift the keyboard over to put the numpad under your left hand, leaving your right hand free to do whatever with the mouse. Of course these days, it seems companies are more interested in rearranging the pageup/down cluster than doing something innovative.
A couple of years ago, Logitech put out their DiNovo with a detachable wireless keypad. You could use it on whatever side you wanted. You could use it as a four-function calculator. You could use it as a media remote. You could reprogram it to do whatever you wanted. Sadly, the keys themselves were mushy crap, and not very enjoyable to type on.
There was a time when a gesture plugin was the first thing I installed when installing a copy of Firefox. Now, I really rarely use them, and would really favor chording and additional buttons. Traditionally, your left button is for clicking, and right button is to initiate a gesture. Lets say you want to go forward or backward, a right click followed by dragging the mouse is a very deliberate action and takes time. Less time than going up to the forward/back button, but considerably more effort than a chord such as left-hold-right, or right-hold-left. They all take more time than just having a forward and back button on your mouse.
For opening a tab, you could have different gestures for foreground and background, or you could do something like middle click for foreground, right-hold-middle for background. A middle click somewhere other than a link could close the page. Right-hold-forward/backward could be used to cycle through a list of favorites. We've got five fingers, and without a whole lot of effort, muscle memory allows you hit multiple buttons each without thinking about it. Why limit yourself to a single button that does everything. Put all your fingers to work. Think of it like the difference between hunt-and-peck and touch typing.
There's no reason why you couldn't install multiple power generation systems in such a plant. You can run a gas turbine engine on just about any heat source, it doesn't have to be combustible fuel. Solar powered gas turbines have been tested and run at comparable thermal efficiency to steam turbines, with no water requirement. For daytime use, run the gas turbine, and deflect some of the energy to heat up your molten sodium tank. During the night, use the energy stored in the sodium tank with a separate steam turbine. You could even run both turbines into a single generator to defer much of the cost.
Slashdot Mirror
According to that list, the worst quake they've suffered in hundreds of years is a 7.2. Sure, the Haiti quake caused widespread damage at only a 7.0, but that's because they're dirt poor and living in shacks. With proper preparation, a 7.2 really isn't bad at all. Seismic isolation foundations have been in use for over two millennia. Building such structures on the containment building and reactor core would make it easily capable of handling such a quake.
No, but if you have no pre-existing union contract ensuring otherwise, they can fire you and hire replacements when you refuse to come to work.
So it's the old trick where an object in a glass of water is not where it appears to be?
I'm surprised no one has yet commented to this effect, but why would you want to use this patent? As I read it, the patent is for a very simple feedback control system for positioning of heliostats (mirrors). You put a camera on the collector, pointed at the mirror, and the camera controls the alignment of the mirror to center the point of highest intensity (the sun). Seems simple enough.
The first problem, this only works for a single mirror. That means you would need one of these light intensity sensors for each individual mirror, of which there may number thousands. Each of these is going to be on the collector, potentially blocking a significant amount of light from those mirrors. Now you could put a single camera on a rotating boom, allowing it to move around and individually manage each mirror in sequence, but that's still an overly complicated system.
You know the layout of your plant, or at least you should. Why not just use a single camera, tracking the sun across the sky, and use that combined with a bit of geometry to determine the optimum placement of each mirror to follow it. The other system has the advantage of being able to track the source of highest intensity, but surely any other source of light will be inconsequential compared to the sun. The next closest object (the moon) at its brightest might only provide a few kW of power to a several hundred MW plant.
But wait! There's more! The sun is a celestial object, and celestial objects are nothing if not predictable. Why bother with cameras at all? A nominal amount of CPU power would be able to predict the sun's track across the sky with micro-arcsecond accuracy. There's absolutely no need for any sort of feedback system at all, besides the position sensors built into the servo motors themselves. This just seems like Google had some image processing expertise, and decided to throw science at the wall to see what would stick.
This isn't a voice/video chat specification. This is a framework for direct browser to browser communication, using a central server as a negotiator. Google Talk (jabber) would then be built on top of this framework.
The issue is that it requires flash, and is only implemented in flash. If you want to use it, you have to write your application in flash. Similarly, Skype only works with the Skype client, and if you want to use their P2P voice network, you have to use their software. The difference here is that its an open specification. Anyone is free to implement it however they choose, and are not constrained to one company's support of a closed binary application.
a microwave could be up to 1300 watts and the full force of one will cook you quite quickly).
Oh come on. Even in open air, with no shielding, the power density of a 1.3kW magnetron falls below that of normal sunlight at somewhere under two feet.
Wait. I was under the impression that the GM plants were engineered to be incapable of pollination, to require farmers to purchase new seeds every crop. Clearly if they are pollinating to other fields, the plant in question is no longer the patented technology in question
Actually they make gas turbine engines for (expensive) RC aircraft that put out between 50-100 lbs of thrust, or you could build your own out of a high end automotive turbocharger. The issue is that you're right back to the big ducted fan, trading a largish two-cycle reciprocating engine for a smaller turbine. The OP was talking about a jetpack of similar size to a hikers pack, which would mean you would be using those RC turbines directly.
Actually ducted fans are jets. The term jet simply refers to an entrained flow inside a surrounding medium. Ducted fans produce jets with a rotor in a cowling. Pumpjets produce jets with an impeller in a cowling. Rivers produce jets as they flow into the ocean. Your garden hose produces a jet when you open the valve.
A turbojet is one specific type of jet produced by the exhaust of a gas turbine engine. A turbofan is another type of jet produced by shaft output of a gas turbine engine powering a ducted fan. There is nothing inherently tying a gas turbine engine to a jet.
it could be as much as 38% efficient
Two stroke engines don't run anything close to that efficiency, and the more highly stressed they are, the more unburned oil you mix into the fuel for lubrication.
A rocket is a type of jet.
Those two spinning rotors in a duct, those are called jets.
It straps onto your back, and it achieves all its lift using a jet. Sounds like a jetpack to me.
Strap a couple of those miniature jet engines (you'll probably need four for your weight, their weight, and a decent amount of fuel) to your back side, and if you don't burn your legs off, you'll at least start a brush fire. As simple as you make the concept out to be, the reason no one has ever bothered to do that is because no one has figured out a way to resolve the problem of exhaust temperature.
Meanwhile a large ducted fan means significantly lesser power requirement, and similarly, significantly less waste heat you have to deal with.
There are three factors you have to consider when building something like this: thrust to weight, thrust to power, and jet temperature. Your thrust to weight needs to be above unity, for obvious reasons. The most efficient way to achieve that thrust is a very large fan, at a very low flow velocity. As you try to shrink the unit, your fan becomes smaller, you need higher and higher flow velocity to achieve the same thrust. As you trade size for power, you go from reciprocating ducted fan, to high bypass turbofan, to low bypass turbofan, to turbojet. Each step means reduced efficiency and significantly higher exit temperature.
What you ask is certainly do-able with a turbojet, however your exhaust would start brush fires, melt asphalt, burn off your legs, and depending on how long you stayed near the ground, suffocate you and eventually itself. Now before you say 'but but but... current jetpacks are that small', that's because current jetpacks are all cold gas rockets. They have a tank of hydrogen peroxide and a catalyst. The catalyst causes rapid disassociation, and the resultant jet is a mixture of water vapor and oxygen. There is no combustion to generate heat, and as a result, they are ridiculously inefficient, which is why you get operation times of under a minute.
Yes. If you get up high enough that icing might be a problem, you pass out yourself from lack of oxygen.
Alpha and beta radiation is radiation for as long as it is actually radiating. As soon as it impacts a surface and sticks, it becomes helium and electrons.
Radiation is short lived, and not a contaminate you can simply remove. Isotopes undergoing decay to produce said radiation can be removed.
Certainly with the PS3, the Cell was not an off-the-shelf design, but was an entirely new architecture, requiring entirely new compiler research. Now with the PS4, the architecture is stable, the compiler research has been done. Process migration, and duplication for multi-core, carries its own difficulties, but those are relatively simple compared to the initial development. It shouldn't take nearly the amount of time scaling the chip for 4-8 times the performance of the original Cell, if by no other means than by adding more parallelism. Sony funded the Cell expecting it would run their whole multimedia lineup for the next 20 years. I don't expect them to abandon all those costs and switch to a new processor.
While there are certainly production limits with XDR, graphics memory is plentiful and cheap. GDDR5 would work just fine as a replacement high speed store on a next generation Cell, of course there would be a fair bit of development spent replacing the existing memory controller.
Off-the-shelf CPUs/GPUs, or custom ASICs using 3rd-party licensed CPU/GPU designs (instead of designing one from scratch)
The last two generations of consoles have used off-the-shelf GPUs. At most, they have the manufacturer bring in some new tricks bound for their next generation as a boost. They do not design one from scratch, and there is no reason for them to consider doing so in the future.
The Cell has already been designed. It is now an off-the-shelf processor. There is no need to design a new architecture from scratch. Take the original Cell, increase size up the ring bus, double the SPEs, and then put 4-6 of those modules on a single die with cache and a shared memory controller. By the time the PS4 is released, IBM will be on a 22nm production process, and the new chip will actually be smaller, cheaper, and use less power than the original.
Off-the-shelf DDR(1/2/3/4/5/whatever) SDRAM (instead of using something from Rambus)
I only partially agree with this one. With the kind of floating point performance the Cell was offering, standard DDR2 at the time simply didn't cut it. DDR3 didn't support those kinds of speeds until over two years after the release of the PS3. As it stands, the Cell is pretty heavily tied into Rambus interfaces, and I don't know if that could be worked around. IMHO, they should use ~1GB of XDR2 or some form of GDDR at 100GB/s+, for the same reasons as before. Then tack on another 4-8GB of cheap DDR3. It's a travesty that the current consoles are stuck with so little memory. Optical mediums are slow. Rotating disks are slow. Use a shit load of memory and use it as a data prefetch to get rid of any form of loading time. Dump anything you need in there for the next couple minutes, and then use the high speed GDDR/XDR as a low speed cache.
Blu-ray, instead of a new kind of optical disk design (or, even eliminate the physical medium altogether in favor of online purchases)
The new optical format is already designed and pushed into mass market. The most they might want to consider would be switching to a 4-layer design, which really doesn't cost a whole lot.
Sony has done the same thing with the PS3. There have been about a dozen different hardware revisions since the original two release units, with multiple die shrinks on the silicon. Production costs are probably under $200, and power consumption is about a third what it originally was.
The problem is that the keyboard is designed for typing, not command input, and the existing chords are not well laid out. For instance, you use ctrl+tab to cycle through your tabs. Now you want to close one. With the mouse gesture, it's a click and a flick of the wrist to close it. It takes moving your whole arm to contort your hand into the shape needed to hit ctrl+f4. Now a Nostromo or similar programmable keypad would work great in this regard, but then you're no longer talking about controls that are standard to an OS or UI paradigm.
Personally, I don't understand why the number pad didn't shift to the left side of the keyboard for exactly this reason 20 years ago when the mouse started becoming popular. Unless you enjoy one-handed slow typing, you've got to move your hand over all that dead space to get it onto the character section. Conversely, if you actually do want to use the numpad for control or number entry, you either have to use your right hand, leaving your left hand is sitting idle and useless, or you need to have a foot and a half extra space to shift the keyboard over to put the numpad under your left hand, leaving your right hand free to do whatever with the mouse. Of course these days, it seems companies are more interested in rearranging the pageup/down cluster than doing something innovative.
A couple of years ago, Logitech put out their DiNovo with a detachable wireless keypad. You could use it on whatever side you wanted. You could use it as a four-function calculator. You could use it as a media remote. You could reprogram it to do whatever you wanted. Sadly, the keys themselves were mushy crap, and not very enjoyable to type on.
Small flaws like how they don't scale beyond small, handheld devices?
There was a time when a gesture plugin was the first thing I installed when installing a copy of Firefox. Now, I really rarely use them, and would really favor chording and additional buttons. Traditionally, your left button is for clicking, and right button is to initiate a gesture. Lets say you want to go forward or backward, a right click followed by dragging the mouse is a very deliberate action and takes time. Less time than going up to the forward/back button, but considerably more effort than a chord such as left-hold-right, or right-hold-left. They all take more time than just having a forward and back button on your mouse.
For opening a tab, you could have different gestures for foreground and background, or you could do something like middle click for foreground, right-hold-middle for background. A middle click somewhere other than a link could close the page. Right-hold-forward/backward could be used to cycle through a list of favorites. We've got five fingers, and without a whole lot of effort, muscle memory allows you hit multiple buttons each without thinking about it. Why limit yourself to a single button that does everything. Put all your fingers to work. Think of it like the difference between hunt-and-peck and touch typing.
There's no reason why you couldn't install multiple power generation systems in such a plant. You can run a gas turbine engine on just about any heat source, it doesn't have to be combustible fuel. Solar powered gas turbines have been tested and run at comparable thermal efficiency to steam turbines, with no water requirement. For daytime use, run the gas turbine, and deflect some of the energy to heat up your molten sodium tank. During the night, use the energy stored in the sodium tank with a separate steam turbine. You could even run both turbines into a single generator to defer much of the cost.