Insane, OK, but there is a limit to information compression. Believe me, all of these tricks are being used to their max and it's still a problem. There is no exponential increase in compression efficiency. If you take compression, faster disks and RAM, and improvements in networking technology into consideration, you still have a huge gap arising between CPUs and storage/memory in the fairly near future.
Those are indeed memory and disk improvements, but they are not keeping up with Moore's law the way CPU speeds are. So there is an increasing gap. I am one who believes that the gap must ultimately be closed, but right now, it is not being paid much attention because FLOPS are still king marketing-wise.
Yes, it's a trick to see if you know what you are talking about. You see, CPUs are scaling by Moore's law, which is exponential. Disk I/O is not scaling by an exponential. The consequences are left as an exercise for the reader.:-) Do you know how many disks you are going to need to keep up with the CPU? The answer is a number that costs way more than the machine itself. So, yes, it can be done, but the cost is going to be more and more about I/O. In fact, that relative part of the cost will increase exponentially. Economics being what they are, this will make exascale computers very very expensive, which means they won't get built.
No, that won't work. I'm talking about persistent storage. What do you do when you want to view the results of timestep 45 in a 1000 timestep simulation, each timestep of which generates a few petabytes of data?
the big problem is once you solve an exascale problem, how do you write the answer to disk fast enough? Petabytes of data are hard to handle/visualize/analyze/copy around.
Yes, exascale is a word. No, exascale does not mean "very large." It means "1000 times as big as petascale." You might want to check Google before you post. I know, this is/.
They only were able to take down Al Capone after charging him with tax evasion.
That is not an argument for using legal dirty tricks that endanger the innocent of false accusations, it's an argument for determining why the obviously guilty cannot be punished.
My link does not "prove me wrong," you hairbrained twit. It shows that we firebombed that city and killed many innocent people. I agree we sent them leaflets first, how nice of us. Then we dropped bombs on them because they had the audacity to ignore our leaflets. This killed them, which was intentional on our part. We are responsible for that. Your argument is that if I point a gun at you and tell you to move aside and you don't, then you get what's coming to you. Sure, maybe you're an idiot for not movie, but if I pull the trigger, then I murdered you. Now, we now must talk about whether my murder of you was justified, but it certainly was an intentional murder, right?
The Japanese are not criticizing us, I don't think. I'm calling the practice of killing innocents in warfare into question. If you think it's legitimate for us to firebomb the Japanese women and children to end the war quickly, then it's hard to argue persuasively that others should not bomb civilians for their own purposes, which seem equally compelling to them. At any rate, it's an important question, whichever way you believe.
I love your wikileaks comment. I am in favor of all leaks, revelations, and dissemination of truthful information. Yes, it means my privacy is shrinking, but as long as my leaders' privacy shrinks commensurately it's a net benefit. Cockroaches like to hide in dark crevices.
I gather from your post that you believe that military personnel targeting foreign civilians during open hostilities is a moral and justified act? That we should not pay lip service to the notion that civilians should not be targets?
I read the link. Nothing there disputes my assertion, it simply establishes that our bombings there killed hundreds of thousands of civilians. You know, even if you drop leaflets asking people to leave their homes (and go where exactly?), when push comes to shove you are still at best knowingly killing hundreds of thousands of innocents as "collateral damage" or worse, "means to an end." In fact, however, the cities were carpet bombed and the civilian casualties seemingly were intentional, they were means to an end, to "break the will of the enemy." Through terror. Well, I guess war is hell.
I don't know, but it's not relevant, because you know what, they weren't given that choice, and to them both would have sucked at the time. In retrospect, I'm sure they'd choose the US, who wouldn't? Russia is almost a failed state. I'm not saying we are a bad country, whatever that means, we have a lot of freedoms and greatness in our culture and our history. One of our national principles I think is not to murder innocent people. But sometimes we have done exactly that, and there has been no opportunity to review those actions in a clear fashion. We need to be cognizant of that fact and own up to it or we will lose credibility with the world and with ourselves.
Little talked-about fact: Hiroshima was targeted to maximize civilian casualties. Our government said it was to take out factories, but we intentionally aimed at civilians, killing hundreds of thousands of noncombatants. It was not a new tactic -- before then, we aimed at civilians with firebombs. Terrorism is only terrorism when non-state actors do it, I guess! Yes, I know I'm off-topic.
Hmm, when you say you were maxed out on CPU with a game -- did you mean that your graphics card was maxed out? Are games really that CPU intensive? But it's a fair point, that increasing FLOPS/Watt is an advantage for small portable devices. I am guessing that your GPU (not CPU) was maxed out and that more memory in your GPU would make a big difference.
Well, actually there are two possibilities; either you have a single-core machine, in which case the current conversation doesn't really apply to you as you are not taking advantage Moore's law anyhow, or Firefox is not making use of all that horsepower. Put good software on a good machine and what will limit it is memory and/or I/O, if anything.
Slashdot Mirror
It's not just on that one paper. Google it. E.g. http://www.google.com/hostednews/afp/article/ALeqM5g6udzhonWNxb7k0JA5m3zUajYjqA?docId=CNG.b8e9e74ca1623255c8dd0b5233b5ac93.121
Yes, cooling costs on a large server are substantial -- you have to run huge cooling towers to cool them down.
Direct democracy is a mistake.
The solution to this is called Instant Runoff Voting. Here is their incredibly lame website, which explains the details.
Insane, OK, but there is a limit to information compression. Believe me, all of these tricks are being used to their max and it's still a problem. There is no exponential increase in compression efficiency. If you take compression, faster disks and RAM, and improvements in networking technology into consideration, you still have a huge gap arising between CPUs and storage/memory in the fairly near future.
Those are indeed memory and disk improvements, but they are not keeping up with Moore's law the way CPU speeds are. So there is an increasing gap. I am one who believes that the gap must ultimately be closed, but right now, it is not being paid much attention because FLOPS are still king marketing-wise.
Yes, it's a trick to see if you know what you are talking about. You see, CPUs are scaling by Moore's law, which is exponential. Disk I/O is not scaling by an exponential. The consequences are left as an exercise for the reader. :-) Do you know how many disks you are going to need to keep up with the CPU? The answer is a number that costs way more than the machine itself. So, yes, it can be done, but the cost is going to be more and more about I/O. In fact, that relative part of the cost will increase exponentially. Economics being what they are, this will make exascale computers very very expensive, which means they won't get built.
You and I have a different view of the meaning of the word "working."
No, that won't work. I'm talking about persistent storage. What do you do when you want to view the results of timestep 45 in a 1000 timestep simulation, each timestep of which generates a few petabytes of data?
the big problem is once you solve an exascale problem, how do you write the answer to disk fast enough? Petabytes of data are hard to handle/visualize/analyze/copy around.
Yes, exascale is a word. No, exascale does not mean "very large." It means "1000 times as big as petascale." You might want to check Google before you post. I know, this is /.
Relevance to the topic at hand? You must be new here. :-)
it is a slippery grey area
Ewww!
They only were able to take down Al Capone after charging him with tax evasion.
That is not an argument for using legal dirty tricks that endanger the innocent of false accusations, it's an argument for determining why the obviously guilty cannot be punished.
Yes, I spent two years in Korea and the brutality of the Japanese towards the Koreans was appalling.
My link does not "prove me wrong," you hairbrained twit. It shows that we firebombed that city and killed many innocent people. I agree we sent them leaflets first, how nice of us. Then we dropped bombs on them because they had the audacity to ignore our leaflets. This killed them, which was intentional on our part. We are responsible for that. Your argument is that if I point a gun at you and tell you to move aside and you don't, then you get what's coming to you. Sure, maybe you're an idiot for not movie, but if I pull the trigger, then I murdered you. Now, we now must talk about whether my murder of you was justified, but it certainly was an intentional murder, right?
there were more people killed in a single firebomb raid on Tokyo than in BOTH atomic bombings combined
I doubt if the victims cared whether it was an atomic bomb or another bomb that killed them. Either way we targeted civilians, whatever the reason.
The Japanese are not criticizing us, I don't think. I'm calling the practice of killing innocents in warfare into question. If you think it's legitimate for us to firebomb the Japanese women and children to end the war quickly, then it's hard to argue persuasively that others should not bomb civilians for their own purposes, which seem equally compelling to them. At any rate, it's an important question, whichever way you believe.
I love your wikileaks comment. I am in favor of all leaks, revelations, and dissemination of truthful information. Yes, it means my privacy is shrinking, but as long as my leaders' privacy shrinks commensurately it's a net benefit. Cockroaches like to hide in dark crevices.
I gather from your post that you believe that military personnel targeting foreign civilians during open hostilities is a moral and justified act? That we should not pay lip service to the notion that civilians should not be targets?
I read the link. Nothing there disputes my assertion, it simply establishes that our bombings there killed hundreds of thousands of civilians. You know, even if you drop leaflets asking people to leave their homes (and go where exactly?), when push comes to shove you are still at best knowingly killing hundreds of thousands of innocents as "collateral damage" or worse, "means to an end." In fact, however, the cities were carpet bombed and the civilian casualties seemingly were intentional, they were means to an end, to "break the will of the enemy." Through terror. Well, I guess war is hell.
I don't know, but it's not relevant, because you know what, they weren't given that choice, and to them both would have sucked at the time. In retrospect, I'm sure they'd choose the US, who wouldn't? Russia is almost a failed state. I'm not saying we are a bad country, whatever that means, we have a lot of freedoms and greatness in our culture and our history. One of our national principles I think is not to murder innocent people. But sometimes we have done exactly that, and there has been no opportunity to review those actions in a clear fashion. We need to be cognizant of that fact and own up to it or we will lose credibility with the world and with ourselves.
Little talked-about fact: Hiroshima was targeted to maximize civilian casualties. Our government said it was to take out factories, but we intentionally aimed at civilians, killing hundreds of thousands of noncombatants. It was not a new tactic -- before then, we aimed at civilians with firebombs. Terrorism is only terrorism when non-state actors do it, I guess! Yes, I know I'm off-topic.
Hmm, when you say you were maxed out on CPU with a game -- did you mean that your graphics card was maxed out? Are games really that CPU intensive? But it's a fair point, that increasing FLOPS/Watt is an advantage for small portable devices. I am guessing that your GPU (not CPU) was maxed out and that more memory in your GPU would make a big difference.
Well, actually there are two possibilities; either you have a single-core machine, in which case the current conversation doesn't really apply to you as you are not taking advantage Moore's law anyhow, or Firefox is not making use of all that horsepower. Put good software on a good machine and what will limit it is memory and/or I/O, if anything.