Yes. But taking the engine out of one car and putting it into a completely redesigned chassis is analogous to taking the GPU pipeline apart and putting it back together in a way that no longer resembles a GPU pipeline.
in theory yes. in practice, it is a lot more complicated then that. GPU's are very good at crunching floating point style operations on wide data sets and have very wide and deep pipelines that are heavily optimized for that purpose. their pipelines are not well suited for the general purpose computing problems that an OS has to deal with which are riddled with branches and interrupts. without a branch predictor, the GPU pipeline would have to stall on every branch and if it had a branch predictor the branch mispredict penalty would be atrocious because the pipelines are so deep.
Slapping the trappings needed to support all those nitty gritty OS-level details onto a GPU pipeline would suit no useful purpose. It would be like taking the trailer off the cab of an 18 wheeler and hitching it to the back of a ferrari. Yes, its theoretically doable, but none of the ferrari's strengths are suited for the purpose of hauling freight and the end result would be comical.
I think the fact that nvidia is talking about coupling x86 cores with a GPU proves that they agree. Packaging them together into a single die for low-power applications makes sense. Forcing them together even further so the GPU pipeline is also handling the general purpose side of the computing needs does not.
I suspect that what you are demonstrating is only that the "no hair theorem" is not a statement about unstable black hole solutions to the gravitational field equations which have not yet collapsed into a stable solution. It is a statement about gravitationally stable solutions to those equations.
I don't believe you are understanding what the no hair theorem really says.
The no hair theorem only refers to parameters of the blackhole that are independent of the frame of reference. A gravitational wave generated by a accelerating mass isn't independent from the frame of reference. Different frames will view it differently.
Also, the poster who said the gravitational wave says something about what is going on inside the black hole is incorrect I believe. The gravitational wave says something about the black hole's velocity and acceleration through space-time because it is that acceleration which produces the waves. The wave says nothing about what may or may not be occurring on the inside.
It wasn't ignored only because the prevailing theory said it can't be right. It was also ignored because there was no possible way to reproduce the experiment. When experimental results go against established theory, science usually gets to the bottom of it by redoing the experiment multiple times over to get stronger confirmation. One experiment unfortunately, all by itself, is not useful. It needs to be reproducible. The fact that this one was not reproducible was the problem.
I too would like to see the CISC instruction sets of our desktop PC's replaced with a simpler RISC architecture. At the same time though I am beginning to realize that the CISC/RISC question is one of aesthetics. Nobody has ever really proven that it makes any difference, and the long standing success of the x86 architecture (not even Intel was able to kill it [...Itanium...]) suggest maybe it really doesn't matter.
Intel closed some very old fabs that were using process technologies that were very out-dated. Those FAB closures were all over the globe, not in the US. As other has said, that was inevitable. The economic downturn hastened it a bit, but it was going to happen anyway. That is always what happens to really old fabs.
But another thing that Intel has announced that it is doing at the same time is spend 7 billion dollars upgrading their state of the art FABs in Oregon, Arizon, and New Mexico; the fabs that manufacture the profitable high-end chips using the latest process technologies.
They hastened the closure of the older fabs that were in the process of closing down anyway, but appear to be doubling down in investment in their high end fabs rather than closing anything important.
This deal with TSMC isn't about fab closure at all. It's about how they want to strategically position the atom processor relative to the competition in the low margin market segments.
I found nothing there that proves your point. every endeavor has regulations. how does pointing at the regulation in this area demonstrate your point? I am aware of private companies that are already doing the things which you say they cannot do.
you may be right. though the difference here is that the Star Trek trademark was and is still in active use by Paramount. All products using the netbook trademark on the other hand were discontinued it seems before any one else started using the term.
that is what the article says the proposal will be: send the robots up early, have them scope out some potential sites and figure out from there which approach makes the most sense: the berms, or the paving approach.
nobody is stopping private industry from competing. private industry is in fact beginning to catch up and will supplant NASA if it ever does. it has a ways to go yet.
you are forgetting that it is enormously expensive to export human labor to the moon because humans require atmosphere to prevent their blood vessels from exploding and require oxygen to prevent suffocation. humans also require that the landing on the surface be a light touch down to prevent them from being crushed on impact.
robots, have none of these issues. they don't breath. they can be designed to function in the absence of atmosphere and they can be built to withstand high impact landings (which are cheaper) that would crush a human.
none of those are issues on earth because earth has an atmosphere, that atmosphere is made up of breathable air, and driving or taking public transportation to a work site is much simpler and cheaper then sitting on the head of a giant missile.
This is true. Intel and the other companies currently using the term netbook to define this class of devices could have said "Ooops sorry" and that would have been the end of it.
If Psion had made their move to protect the trademark back when the Intel first started encouraging the use of this term in the media then that surely is what would have happened. But psion, in their own statements, said they waited until the latter quarters of 2008 after it was clear to them that the use of the term netbook for this general class of devices was spreading across the industry.
That I think is a key distinction. They waited. I'm not a lawyer of course, but it seems relevant to me.
Of course if Psion had sent out the CDC orders sooner, before the term caught on in the media and in the marketting departments of Intel, Dell, and others, then they would have nothing to gain. Psion had already discontinued their netbook line and would have gained nothing if they stopped the industry from using the term before it caught on.
Now look at the position Psion is in. If they succeed, they will now have exclusive rights to a trademark which has value because of the dollars spend by the big players in the industry to push the term as the definition of a general class of device.
It's unfortunate that these companies ended up using a term that was also a trademark for a discontinued psion product line. But it seems highly disengenuous that psion waited until after multiple companies had spent large amounts of marketing cash pushing the term to send their CDC orders. Seems like they waited for these other companies to breath life into the dead trademark before choosing to exercise their claim to that trademark.
i did notice the word console in the summary. of course, the second item on their list, tetris, wasn't a console game. It was ported to many consoles long after it was already a very old game. But it was an arcade game from the 80s that wasn't ported to any consoles until much later.
I agree. AMD likely wants generic "platform" solutions (where they provide the customer with multiple components of the platfrom rather than just the CPU) for as many markets as they can hit rather than just the basic business segment. That probably is why they chose ATI as opposed to something cheaper.
exactly. i love my nvidia gpu because I am a gamer and a scientist. but...the thing draws 2X the power of my old core 2 duo, and in the business environment where you have no business need for any of the options that this fancy gpu makes available to you...its an absurd proposition.
try telling your manager that you need a fancy gpu, and a bigger power supply to keep it from blue screening the machine because it draws so much current, all because you want to make fancy animated power point presentations for your co-workers. He'll likely laugh you out of his office.
as I mentioned in another reply elsewhere though, it isn't really about the "average Joe". It's about business customers who buy PC's to put under their employee's desks. THAT is where the vast majority of desktop PC sales actually occur and that is the reason Intel is the leader. Businesses do not need or even want to have their employees using PCs in the office that are capable of playing games well. They want their employees using office apps and running email programs.
It is not about bundling. It is about the fact that the vast majority of PC sales are to business customers who want to put desktops under the desks of their employees and don't give a damn about the GPU performance. To those customers, spending the premium for an nVidia GPU is absurd. Hence, they buy inexpensive machines that have GPU's which suck at rendering 3D but are fully functional when it comes to running Office or Email applications. This, btw, is in my opinion the real reason AMD bought ATI. AMD wanted to work toward having a solution for that high volume market, and seemed to think they needed to own ATI to do it.
Many of the people who put together high end machines for gaming and/or other 3D application purposes---the people that buy and value what nVidia has to offer---frequently forget that type of machines they love are a very tiny percentage of the desktop market...
The article does not state that Vista users are getting free upgrades to windows 7. It says that people who buy new PC's after the upgrade program takes effect will be eligible for a free upgrade to windows 7 if those new PCs came with Vista installed.
That is the exact same upgrade program they have offered in the months preceding the release of every version of the windows operating in recent memory.
Thank you for the link to Mr. Hogan's page. Mr. Hogan may very well be right that electromagnetism may play a larger role than astronomers currently assume.
But he is mistaken about one point. He claims that all current observations of black holes have been observations of highly energetic events that then get attributed to black holes. He is implying that the only observations made so far have been the cases where highly energetic jets coming out of galaxies have been observed.
He is wrong about that.
Yes, those observations frequently are explained as possible black holes. But some of the more convincing observations of apparent black holes have been observations of stars that orbit rapidly around a center of mass distinct from the star as though the star was part of a binary start system in which one of the stars was not visible. In other words, binary star system in which one of the stars in the system emitted no light.
Energetic plumes can potentially be explained away as something other than a black hole, but when you find a star that behaves as though it was part of a binary system but don't see it's neighbor...then run the solution of the gravitational equations given the motions you do observe to determine the location and mass of the "missing" star, and see that the equations show a very high density mass that you know you wouldn't be able to see if it did exist... that is something quite different than what Mr. Hogan is talking about.
Slashdot Mirror
Yes. But taking the engine out of one car and putting it into a completely redesigned chassis is analogous to taking the GPU pipeline apart and putting it back together in a way that no longer resembles a GPU pipeline.
in theory yes. in practice, it is a lot more complicated then that. GPU's are very good at crunching floating point style operations on wide data sets and have very wide and deep pipelines that are heavily optimized for that purpose. their pipelines are not well suited for the general purpose computing problems that an OS has to deal with which are riddled with branches and interrupts. without a branch predictor, the GPU pipeline would have to stall on every branch and if it had a branch predictor the branch mispredict penalty would be atrocious because the pipelines are so deep.
Slapping the trappings needed to support all those nitty gritty OS-level details onto a GPU pipeline would suit no useful purpose. It would be like taking the trailer off the cab of an 18 wheeler and hitching it to the back of a ferrari. Yes, its theoretically doable, but none of the ferrari's strengths are suited for the purpose of hauling freight and the end result would be comical.
I think the fact that nvidia is talking about coupling x86 cores with a GPU proves that they agree. Packaging them together into a single die for low-power applications makes sense. Forcing them together even further so the GPU pipeline is also handling the general purpose side of the computing needs does not.
I suspect that what you are demonstrating is only that the "no hair theorem" is not a statement about unstable black hole solutions to the gravitational field equations which have not yet collapsed into a stable solution. It is a statement about gravitationally stable solutions to those equations.
I don't believe you are understanding what the no hair theorem really says.
The no hair theorem only refers to parameters of the blackhole that are independent of the frame of reference. A gravitational wave generated by a accelerating mass isn't independent from the frame of reference. Different frames will view it differently.
Also, the poster who said the gravitational wave says something about what is going on inside the black hole is incorrect I believe. The gravitational wave says something about the black hole's velocity and acceleration through space-time because it is that acceleration which produces the waves. The wave says nothing about what may or may not be occurring on the inside.
It wasn't ignored only because the prevailing theory said it can't be right. It was also ignored because there was no possible way to reproduce the experiment. When experimental results go against established theory, science usually gets to the bottom of it by redoing the experiment multiple times over to get stronger confirmation. One experiment unfortunately, all by itself, is not useful. It needs to be reproducible. The fact that this one was not reproducible was the problem.
I guess you didn't read any of the articles linked in the post. The company that makes this stuff also makes paper from hemp.
I too would like to see the CISC instruction sets of our desktop PC's replaced with a simpler RISC architecture. At the same time though I am beginning to realize that the CISC/RISC question is one of aesthetics. Nobody has ever really proven that it makes any difference, and the long standing success of the x86 architecture (not even Intel was able to kill it [...Itanium...]) suggest maybe it really doesn't matter.
The ever escalating frequency wars ended about 4 or 5 years ago. Welcome back. From where rock you were hiding under. ;)
Intel closed some very old fabs that were using process technologies that were very out-dated. Those FAB closures were all over the globe, not in the US. As other has said, that was inevitable. The economic downturn hastened it a bit, but it was going to happen anyway. That is always what happens to really old fabs.
But another thing that Intel has announced that it is doing at the same time is spend 7 billion dollars upgrading their state of the art FABs in Oregon, Arizon, and New Mexico; the fabs that manufacture the profitable high-end chips using the latest process technologies.
They hastened the closure of the older fabs that were in the process of closing down anyway, but appear to be doubling down in investment in their high end fabs rather than closing anything important.
This deal with TSMC isn't about fab closure at all. It's about how they want to strategically position the atom processor relative to the competition in the low margin market segments.
I found nothing there that proves your point. every endeavor has regulations. how does pointing at the regulation in this area demonstrate your point? I am aware of private companies that are already doing the things which you say they cannot do.
both of these are private companies: http://www.scaled.com/ http://www.virgingalactic.com/
and then a list of 24 additional companies here on the left sidebar of this page: http://space.xprize.org/ansari-x-prize
you may be right. though the difference here is that the Star Trek trademark was and is still in active use by Paramount. All products using the netbook trademark on the other hand were discontinued it seems before any one else started using the term.
but we'll have to see.
that is what the article says the proposal will be: send the robots up early, have them scope out some potential sites and figure out from there which approach makes the most sense: the berms, or the paving approach.
afterall, his computer is smarter than those robots are likely to be...
nobody is stopping private industry from competing. private industry is in fact beginning to catch up and will supplant NASA if it ever does. it has a ways to go yet.
you are forgetting that it is enormously expensive to export human labor to the moon because humans require atmosphere to prevent their blood vessels from exploding and require oxygen to prevent suffocation. humans also require that the landing on the surface be a light touch down to prevent them from being crushed on impact.
robots, have none of these issues. they don't breath. they can be designed to function in the absence of atmosphere and they can be built to withstand high impact landings (which are cheaper) that would crush a human.
none of those are issues on earth because earth has an atmosphere, that atmosphere is made up of breathable air, and driving or taking public transportation to a work site is much simpler and cheaper then sitting on the head of a giant missile.
This is true. Intel and the other companies currently using the term netbook to define this class of devices could have said "Ooops sorry" and that would have been the end of it.
If Psion had made their move to protect the trademark back when the Intel first started encouraging the use of this term in the media then that surely is what would have happened. But psion, in their own statements, said they waited until the latter quarters of 2008 after it was clear to them that the use of the term netbook for this general class of devices was spreading across the industry.
That I think is a key distinction. They waited. I'm not a lawyer of course, but it seems relevant to me.
Of course if Psion had sent out the CDC orders sooner, before the term caught on in the media and in the marketting departments of Intel, Dell, and others, then they would have nothing to gain. Psion had already discontinued their netbook line and would have gained nothing if they stopped the industry from using the term before it caught on.
Now look at the position Psion is in. If they succeed, they will now have exclusive rights to a trademark which has value because of the dollars spend by the big players in the industry to push the term as the definition of a general class of device.
It's unfortunate that these companies ended up using a term that was also a trademark for a discontinued psion product line. But it seems highly disengenuous that psion waited until after multiple companies had spent large amounts of marketing cash pushing the term to send their CDC orders. Seems like they waited for these other companies to breath life into the dead trademark before choosing to exercise their claim to that trademark.
i did notice the word console in the summary. of course, the second item on their list, tetris, wasn't a console game. It was ported to many consoles long after it was already a very old game. But it was an arcade game from the 80s that wasn't ported to any consoles until much later.
Yeah. I am not sure if they chose poorly or not. But they definitely made a very risky choice. ATI cost them a lot of money.
I agree. AMD likely wants generic "platform" solutions (where they provide the customer with multiple components of the platfrom rather than just the CPU) for as many markets as they can hit rather than just the basic business segment. That probably is why they chose ATI as opposed to something cheaper.
CAD does need a decent card. But that's a very small market too. Most business users are running office apps and email, and that is it.
exactly. i love my nvidia gpu because I am a gamer and a scientist. but...the thing draws 2X the power of my old core 2 duo, and in the business environment where you have no business need for any of the options that this fancy gpu makes available to you...its an absurd proposition.
try telling your manager that you need a fancy gpu, and a bigger power supply to keep it from blue screening the machine because it draws so much current, all because you want to make fancy animated power point presentations for your co-workers. He'll likely laugh you out of his office.
as I mentioned in another reply elsewhere though, it isn't really about the "average Joe". It's about business customers who buy PC's to put under their employee's desks. THAT is where the vast majority of desktop PC sales actually occur and that is the reason Intel is the leader. Businesses do not need or even want to have their employees using PCs in the office that are capable of playing games well. They want their employees using office apps and running email programs.
It is not about bundling. It is about the fact that the vast majority of PC sales are to business customers who want to put desktops under the desks of their employees and don't give a damn about the GPU performance. To those customers, spending the premium for an nVidia GPU is absurd. Hence, they buy inexpensive machines that have GPU's which suck at rendering 3D but are fully functional when it comes to running Office or Email applications. This, btw, is in my opinion the real reason AMD bought ATI. AMD wanted to work toward having a solution for that high volume market, and seemed to think they needed to own ATI to do it.
Many of the people who put together high end machines for gaming and/or other 3D application purposes---the people that buy and value what nVidia has to offer---frequently forget that type of machines they love are a very tiny percentage of the desktop market...
Most everything in the summary is wrong. :(
The article does not state that Vista users are getting free upgrades to windows 7. It says that people who buy new PC's after the upgrade program takes effect will be eligible for a free upgrade to windows 7 if those new PCs came with Vista installed.
That is the exact same upgrade program they have offered in the months preceding the release of every version of the windows operating in recent memory.
There is no news here.
Thank you for the link to Mr. Hogan's page. Mr. Hogan may very well be right that electromagnetism may play a larger role than astronomers currently assume.
But he is mistaken about one point. He claims that all current observations of black holes have been observations of highly energetic events that then get attributed to black holes. He is implying that the only observations made so far have been the cases where highly energetic jets coming out of galaxies have been observed.
He is wrong about that.
Yes, those observations frequently are explained as possible black holes. But some of the more convincing observations of apparent black holes have been observations of stars that orbit rapidly around a center of mass distinct from the star as though the star was part of a binary start system in which one of the stars was not visible. In other words, binary star system in which one of the stars in the system emitted no light.
Energetic plumes can potentially be explained away as something other than a black hole, but when you find a star that behaves as though it was part of a binary system but don't see it's neighbor...then run the solution of the gravitational equations given the motions you do observe to determine the location and mass of the "missing" star, and see that the equations show a very high density mass that you know you wouldn't be able to see if it did exist... that is something quite different than what Mr. Hogan is talking about.