You can go RTFA yourself. I read TFA and TFA says "A single GX2 plugs into one PCI Express slot, but it actually has a pair of printed circuit boards". Last time I checked, a "pair" means two. From what I can tell, they have basically just screwed two cards together in a SLI configuration and made them share the PCI slot.
With the current trend of multiple cores, I figured it would be just a matter of time for the SLI and Crossfire solutions to switch back to a single video card. Either they would dual core the GPU, or simply put two GPU on the same card.
Actually, the G71 processor used in that beast has 32 pixel pipelines already, which in their context are similar to cores on a CPU. (Sure, they form a SIMD architecture unlike CPU cores, but so does SLIed GPUs sort of as I have understood it.) When CPUs get more cores, GPUs get more pipelines. I suppose you could put 64 in there instead, but then you run into manufacturing problems because of the chip size.
They could probably put two GPUs on the same board though.
Re:Non-U.S.'ers not safe either
on
Death By DMCA
·
· Score: 2, Interesting
nd once you get a fresh new P2P client to get very popular and using encryption with onion routing, I think that's the final nail in the coffin against **AA's "shut down" or even lawsuit strategy.
Their response will be to try to outlaw that particular client and clients like it, and impose DRM technology on every user to enforce the law. How well that turns out might be a decisive moment for the future of culture, I suspect.
Re:more proof the RIAA/MPAA are insane
on
Death By DMCA
·
· Score: 5, Insightful
Look at it this way - for the people working in those companies, it is their job to get you to watch TV and more specifically watch those ads. They will pursue all means that they think are ethical/legal and probably some that they don't. But it's their job, and you can hardly blame them for doing it
I beg to differ. You can blame them, and in fact you should blame them. That is how a market economy works: if I don't like that a certain shoe manufacturer profits from child labour, then I blame them for it, and stop buying their shoes.
When we accept questionable and dishonest behaviour from corporations, simply because it is somehow expected of them, then that is how they will behave. The truth is, it is expected from them only because we accept it. If we didn't, it would no longer be profitable and they wouldn't do it. Companies have no intrinsic moral; their only moral stems solely from the criticism we as consumers place on them. Humans are the only source of moral in the system, and we must use it.
I read it, although not too carefully (evidently), and it was a while ago. And it didn't seem to contain the word "codec" when I scanned it. But ok, point taken. =P
Yes, you could probably do all that without downloading.debs - agreed. I read it more as an interesting anecdote rather than a tutorial, though.
I posted a link to an article titled "Ubuntu for your grandmother.", as a confirming example of the GGGGP's conjecture. It had nothing to do with codecs,.debs or beer.
It's rather funny how the Airbus people keep restating the obvious.
Well, I wouldn't know about that since I am not an Airbus person, nor know of any. But hey, this is slashdot and I didn't automatically join in on the unison french-bashing chorus, so certainly I have to be biased somehow.
It's called the "shut off the autopilot and FLY" system [...] Patented, of course
As a side not: So they basically renamed an "off" button and managed to get it patented? It is fascinating how messed up the patent system seems to have become, although it has since long ceased to surprise me.
You seem to know more about aviation than I do, so I'm not going to dispute you when you say the pilot is a safer bet than the system. However, I do think that we will eventually reach a point where the failure rate of an automated system is lower than the rate of human error -- a point when routing the decision making of some problems through the pilot actually increases the rate of an error. And I don't think this is too far into the future.
Well, obviously, there would be a way to abort the climbing action. Turning off the autopilot, for example. You could even have a giant red "abort" button if that would help. But the key point is that unless the pilot takes some active action, the aircraft should automatically do what it regards to be safest by default.
IIRC Boing 737:s have done similar things (auto-landed in forrests etc. and not in test flights). So it's not just Airbus. But on the other hand, there are also shitloads of examples where the pilot screwed up. The auto-pilot doesn't have to be perfect, it just has to save the aircraft more often than it induces a crash.
B) You better do what the box says, or you something bad will happen.
If the general idea is to have the the aircraft shout "CLIMB CLIMB" at the pilot, which in turn have "better do what the box says", I'd much rather have the aircraft just climb.
That comment was not only funny, but in a way actually rather insightful -- objects interact with the environment in a lot more ways than through exchange of visible light.
If I were to find a (moving) tank where I live, I would rather go for all the trees and bushes cracking and falling around it as it paves its way through the forrest.
Actually, I have never (unintentionally) gotten any of my PCs infected with a computer virus, but thrice I have had the system severely broken by the virus scanner (each time a different brand). I have started to think it is a greater risk to have a virus scanner installed than not to have one, at least for me...
Theoretically, there is no language that is more or less prone to bugs than any other language as understood in Turing Completeness. Without delving too much into this, it simply states that all languages emulate a Turing machine to some degree and therefore should be capable of everything a Turing machine is capable of (although I don't think this says anything about time/space efficiency).
Please do delve into this, because it makes no sense whatsoever to me. Turing Completeness has nothing to do with how likely you are to make a mistake when you write a program (that is, accidentally write another program than you actually intended to, and then fail to notice it). This is trivial to prove: Give me a day or two and I will design a turing complete language that is so confusing that anything you try to do with it will be riddled with bugs.
Sure, all turing complete languages are theoretically able to represent any program (including its bugs) that exists in one of them, but so what? It says nothing about the probability that you mistakenly introduce unintended behaviour in a program. What matters is how closely the language is able to transcribe the mental concepts of the problem at hand, and how easily mistakes you make result in a seemingly correct program. And there is a huge difference between programming languages with regard to this.
My point is that it is an impossible herculean task for the developers to test any application in every state.
You don't need to examine every possible state a program can occupy in order to make formal assertions about its behaviour. Anyone who has taken a course in theoretical computer science knows this. (The study of formal program correctness is called formal methods.) Also, you might find BitC interesting -- a programming language designed to support formal program verification.
If you want to taint your kernel by adding proprietary modules, more power to you, but you can't redistribute the result. Every user has to add the tainted bits in themselves.
What if I distribute the kernel with instructions on how to add add proprietary module? Would that be OK?
What if I then distrubute the kernel with a helper script that downloads the modules when the user runs it?
What if these modules would reside on the same CD as the kernel, and the script simply copies them from a specific directory instead of copying them from a server?
What if I also include a helper script that automatically installs the modules when the user runs it?
And what if this script is a boot script?
But, oh wait, that sounds an awful lot like what kororaa does already...? Where did I cross the line?
Come on now moderators. How can a troll like this be modded insightful?
The Soviet Union was very advanced in several fields of science (especially theoretical physics and mathematics). They were the first to launch a sattelite orbiting Earth (Sputnik 1), first to put a living being in orbit (the dog Laika), first to put a man in space, first dual-manned flight, first space walk, first to land on the moon (with a probe), built the first space station (Salyut 1).
While I agree there is a point in letting everyone have access to all the information rather than just the government or some other organisation, there remains one big problem: "Everyone" does not have the same resources or abilities to make use of the information (e.g. private individuals vs. governments or corporations).
It wouldn't be very useful for me as a private citizen to snoop around at the local grocery store, but imagine what a large corporation with the data mining and pattern recognition technologies of 2050, built with the help of top psychologists, could do to manipulate me if they could monitor everything I did and said 24/7.
Sure, if surveillance was ubiquitous, we would also know that they were doing this. But such knowledge isn't enough to protect us, evidently. Otherwise commercials would hardly work today.
We aren't even nearly at the stage of nanomachines ("grey goop")
I think you are referring to grey goo. (Which is an end-of-the-world scenario involving self-replicating nanomachines running amok, in case someone here didn't know that already.)
Yes, in theory a 4% party can get equal power to two 48% parties, but in reality this does not happen. If the 4% party gets too pushy, the two large parties would simply both stop cooperating and the result would be a reelection.
If you look closer at it, there isn't really any perfect election system. This has actually been proven mathematically, which is quite interesting. There are, however, systems that are better than others. When electing a single candidate out of many, as in a presidential election, a Condorcet method is much fairer than a traditional majority vote. (Something the US should consider, perhaps...?)
If you took the aforementioned martial arts classes, with a proper focus on self defense, a knife should be no problem.
Excuse me, but how many knife fights have you been in to support this claim? Even with years of martial arts training, with focus on self defense, a knife is a problem. Against a bladed weapon, even the most skilled warrior runs a substantial risk of severe or fatal injury. If you have only taken a few dozen classes, then you are in serious trouble.
What would possibly use that many cores? And for any single task, the thing would not be efficient. What exactly is the point of this?
Remember, the human brain has 100 billion cores, each in itself very inefficient, and yet it is pretty powerful.
There are huge amounts of unexploited parallelism in the tasks our computers perform. The problem is mainly that most of the tools we use (programming languages etc.) are very serial in how they describe and handle problems and solutions. This is natural, of course, since it reflects how computers historically have been operating. But there are limits on how much computational performance one can focus into a single computational unit, and we are reaching them. So expect to see more development in this direction.
Slashdot Mirror
You can go RTFA yourself. I read TFA and TFA says "A single GX2 plugs into one PCI Express slot, but it actually has a pair of printed circuit boards". Last time I checked, a "pair" means two. From what I can tell, they have basically just screwed two cards together in a SLI configuration and made them share the PCI slot.
Actually, the G71 processor used in that beast has 32 pixel pipelines already, which in their context are similar to cores on a CPU. (Sure, they form a SIMD architecture unlike CPU cores, but so does SLIed GPUs sort of as I have understood it.) When CPUs get more cores, GPUs get more pipelines. I suppose you could put 64 in there instead, but then you run into manufacturing problems because of the chip size.
They could probably put two GPUs on the same board though.
Their response will be to try to outlaw that particular client and clients like it, and impose DRM technology on every user to enforce the law. How well that turns out might be a decisive moment for the future of culture, I suspect.
I beg to differ. You can blame them, and in fact you should blame them. That is how a market economy works: if I don't like that a certain shoe manufacturer profits from child labour, then I blame them for it, and stop buying their shoes.
When we accept questionable and dishonest behaviour from corporations, simply because it is somehow expected of them, then that is how they will behave. The truth is, it is expected from them only because we accept it. If we didn't, it would no longer be profitable and they wouldn't do it. Companies have no intrinsic moral; their only moral stems solely from the criticism we as consumers place on them. Humans are the only source of moral in the system, and we must use it.
Yes, you could probably do all that without downloading .debs - agreed. I read it more as an interesting anecdote rather than a tutorial, though.
I posted a link to an article titled "Ubuntu for your grandmother.", as a confirming example of the GGGGP's conjecture. It had nothing to do with codecs, .debs or beer.
I have no idea what you just said.
Indeed.
Well, I wouldn't know about that since I am not an Airbus person, nor know of any. But hey, this is slashdot and I didn't automatically join in on the unison french-bashing chorus, so certainly I have to be biased somehow.
As a side not: So they basically renamed an "off" button and managed to get it patented? It is fascinating how messed up the patent system seems to have become, although it has since long ceased to surprise me.
You seem to know more about aviation than I do, so I'm not going to dispute you when you say the pilot is a safer bet than the system. However, I do think that we will eventually reach a point where the failure rate of an automated system is lower than the rate of human error -- a point when routing the decision making of some problems through the pilot actually increases the rate of an error. And I don't think this is too far into the future.
IIRC Boing 737:s have done similar things (auto-landed in forrests etc. and not in test flights). So it's not just Airbus. But on the other hand, there are also shitloads of examples where the pilot screwed up. The auto-pilot doesn't have to be perfect, it just has to save the aircraft more often than it induces a crash.
If the general idea is to have the the aircraft shout "CLIMB CLIMB" at the pilot, which in turn have "better do what the box says", I'd much rather have the aircraft just climb.
Didn't you just do that?
That comment was not only funny, but in a way actually rather insightful -- objects interact with the environment in a lot more ways than through exchange of visible light.
If I were to find a (moving) tank where I live, I would rather go for all the trees and bushes cracking and falling around it as it paves its way through the forrest.
Actually, I have never (unintentionally) gotten any of my PCs infected with a computer virus, but thrice I have had the system severely broken by the virus scanner (each time a different brand). I have started to think it is a greater risk to have a virus scanner installed than not to have one, at least for me...
Please do delve into this, because it makes no sense whatsoever to me. Turing Completeness has nothing to do with how likely you are to make a mistake when you write a program (that is, accidentally write another program than you actually intended to, and then fail to notice it). This is trivial to prove: Give me a day or two and I will design a turing complete language that is so confusing that anything you try to do with it will be riddled with bugs.
Sure, all turing complete languages are theoretically able to represent any program (including its bugs) that exists in one of them, but so what? It says nothing about the probability that you mistakenly introduce unintended behaviour in a program. What matters is how closely the language is able to transcribe the mental concepts of the problem at hand, and how easily mistakes you make result in a seemingly correct program. And there is a huge difference between programming languages with regard to this.
You don't need to examine every possible state a program can occupy in order to make formal assertions about its behaviour. Anyone who has taken a course in theoretical computer science knows this. (The study of formal program correctness is called formal methods.) Also, you might find BitC interesting -- a programming language designed to support formal program verification.
What if I distribute the kernel with instructions on how to add add proprietary module? Would that be OK?
What if I then distrubute the kernel with a helper script that downloads the modules when the user runs it?
What if these modules would reside on the same CD as the kernel, and the script simply copies them from a specific directory instead of copying them from a server?
What if I also include a helper script that automatically installs the modules when the user runs it?
And what if this script is a boot script?
But, oh wait, that sounds an awful lot like what kororaa does already...? Where did I cross the line?
The Soviet Union was very advanced in several fields of science (especially theoretical physics and mathematics). They were the first to launch a sattelite orbiting Earth (Sputnik 1), first to put a living being in orbit (the dog Laika), first to put a man in space, first dual-manned flight, first space walk, first to land on the moon (with a probe), built the first space station (Salyut 1).
Just to name a few things.
It wouldn't be very useful for me as a private citizen to snoop around at the local grocery store, but imagine what a large corporation with the data mining and pattern recognition technologies of 2050, built with the help of top psychologists, could do to manipulate me if they could monitor everything I did and said 24/7.
Sure, if surveillance was ubiquitous, we would also know that they were doing this. But such knowledge isn't enough to protect us, evidently. Otherwise commercials would hardly work today.
No it doesn't. The refractive index of cubic zirconium is 2.176 compared to 2.417 for diamonds.
I think you are referring to grey goo. (Which is an end-of-the-world scenario involving self-replicating nanomachines running amok, in case someone here didn't know that already.)
If you look closer at it, there isn't really any perfect election system. This has actually been proven mathematically, which is quite interesting. There are, however, systems that are better than others. When electing a single candidate out of many, as in a presidential election, a Condorcet method is much fairer than a traditional majority vote. (Something the US should consider, perhaps...?)
It is sort of an inverted DDOS...
Excuse me, but how many knife fights have you been in to support this claim? Even with years of martial arts training, with focus on self defense, a knife is a problem. Against a bladed weapon, even the most skilled warrior runs a substantial risk of severe or fatal injury. If you have only taken a few dozen classes, then you are in serious trouble.
Remember, the human brain has 100 billion cores, each in itself very inefficient, and yet it is pretty powerful.
There are huge amounts of unexploited parallelism in the tasks our computers perform. The problem is mainly that most of the tools we use (programming languages etc.) are very serial in how they describe and handle problems and solutions. This is natural, of course, since it reflects how computers historically have been operating. But there are limits on how much computational performance one can focus into a single computational unit, and we are reaching them. So expect to see more development in this direction.