PhysX lets you execute simulation tasks in your own threads by providing a basic scheduler class interface. So they have no way to force you to use a single thread--just write your code properly. Using the default provided threading model in PhysX is only useful for simplicity because you're too lazy to properly integrate PhysX into your engine's threading model, not performance, and shouldn't be doing it in a pro environment anyway.
You're wrong. The US has been focused on first-strike for years. See Lieber, Keir A.; Press, Daryl G. The Rise of U.S. Nuclear Primacy. Foreign Affairs, March/April 2006, Vol. 85 Issue 2
Not really. The US has long had the unofficial policy of first strike capability while taking minimal return-fire damage. See this peer-reviewed article in a major journal:
Lieber, Keir A.; Press, Daryl G. The Rise of U.S. Nuclear Primacy. Foreign Affairs, March/April 2006, Vol. 85 Issue 2
There cannot be infinite many kernels (whether correct or not) unless the kernels can be of infinite size (because you're dealing with discrete systems).
When people say the ends justify the means, they are forgetting consider that the true ends include any side effects of the means. When you total it up in a fully utilitarian sense, then that statement is meaningless.
If we consider that in the future Western nations wane in power and dominance goes to, oh, say China and/or Russia, will there be not the likelihood of far more suffering and misery than a war that sets back those powers early on?
In 2006 the following paper was published by MIT Press: The End of MAD? The Nuclear Dimension of U.S. Primacy. Keir A. Lieber, Daryl G. Press International Security Spring 2006, Vol. 30, No. 4: 7-44. (pdf here: http://www.mitpressjournals.org/toc/isec/30/4 )
Note that International Security is a top journal on its subject, so this is, in intrawebs speak, "serious business".
The authors argue that over the years US policy has moved from the goal of deterrence by mutually assured destruction towards the, what they argue, US military perceives as a possibility of achieving first strike capability with none or minimal possibility for taking damage from return fire. If you read the paper you will see that their argument is well supported.
While the authors themselves are writing from the point of view that this is a worrisome situation, I end up wondering, given what I wrote at the top of this post, whether this is not in fact a positive development. Despite all the faults of Western nations, I would feel far more comfortable in a world where Western civilization, Western values, and the Western way of life weren't threatened in the future on various fronts. But with a resurgent Russia carrying out technical upgrades to early warning systems and their nuclear arsenal, as well as a China with a very high economic growth, the window of opportunity for the US to take advantage of its nuclear primacy is short. As difficult as it is to say, I would rather millions die now than billions more suffer in the future.
I remember during my undergrad in USF (the Tampa one, not SF) eating some popcorn while sitting on a bench in campus. Two squirrels came up to get some of the popcorn I had dropped on the ground. The problem is that they started fighting over it, surprisingly viciously, right around my legs. One managed to jump on my bare ankle and sink it's teeth--deep enough that I couldn't stop the bleeding and had to go to the campus clinic. The doctor said that in his 15 years working there, I'm the only one to have come after being attacked by squirrels.
I'd say a zoologist like Desmond Morris is a much more reliable reference on cats, and he is in agreement with the grandparent and disagreement with you.
who can't be bothered to go to a polling station to have an effect in an election?
I can only see the minor inconvenience of having to get off your ass in order to vote as a great way to filter out those that obviously don't care enough (and are likely as informed about the issues).
He didn't figure out arbitrary computations, just both addition and multiplication. That other computations can be built on top of those primitives has been known for much longer and is nothing new. So while in a sense your comment is formally correct, it is misleading.
Note that hardware generators are not guaranteed to produce great random numbers either, since they have non-uniform distributions. It's best to mix them up (in a feedback mode) with some other thing that has a good distribution, like a mersenne twister or coupled logistic functions.
Exactly. However, even here there is an implicit assumption--the one-time pad is truly random. If you use a crappy pseudorandom generator, your encryption will be vulnerable.
That is not correct. We're not simply dealing with chunks of encrypted data in a database here. We're talking about actual arithmetic operations. If your system supports at least addition and multiplication on encrypted data, then you can implement any other computation based on these.
One can always use the Intel compiler on Linux. The performance of compiled code is comparable to that the Intel compiler produces with its Windows version.
Slashdot Mirror
PhysX lets you execute simulation tasks in your own threads by providing a basic scheduler class interface. So they have no way to force you to use a single thread--just write your code properly. Using the default provided threading model in PhysX is only useful for simplicity because you're too lazy to properly integrate PhysX into your engine's threading model, not performance, and shouldn't be doing it in a pro environment anyway.
You are incorrect, as usual :)
http://en.wikipedia.org/wiki/Lamarckism#Current_views_on_.22Lamarckism.22
You're wrong. The US has been focused on first-strike for years. See Lieber, Keir A.; Press, Daryl G. The Rise of U.S. Nuclear Primacy. Foreign Affairs, March/April 2006, Vol. 85 Issue 2
"Oh yeah, we could go nuclear - but so can they."
Not really. The US has long had the unofficial policy of first strike capability while taking minimal return-fire damage. See this peer-reviewed article in a major journal: Lieber, Keir A.; Press, Daryl G. The Rise of U.S. Nuclear Primacy. Foreign Affairs, March/April 2006, Vol. 85 Issue 2
Hair contains melanin, which is a semiconductor.
Obviously I'm assuming that accelerating expansion is the correct cosmological model, as the preponderance of evidence implies.
You cannot keep on making larger kernels because you have a finite number of distinguishable quantum states within your Hubble volume.
There cannot be infinite many kernels (whether correct or not) unless the kernels can be of infinite size (because you're dealing with discrete systems).
When people say the ends justify the means, they are forgetting consider that the true ends include any side effects of the means. When you total it up in a fully utilitarian sense, then that statement is meaningless.
If we consider that in the future Western nations wane in power and dominance goes to, oh, say China and/or Russia, will there be not the likelihood of far more suffering and misery than a war that sets back those powers early on?
In 2006 the following paper was published by MIT Press: The End of MAD? The Nuclear Dimension of U.S. Primacy. Keir A. Lieber, Daryl G. Press International Security Spring 2006, Vol. 30, No. 4: 7-44. (pdf here: http://www.mitpressjournals.org/toc/isec/30/4 )
Note that International Security is a top journal on its subject, so this is, in intrawebs speak, "serious business".
The authors argue that over the years US policy has moved from the goal of deterrence by mutually assured destruction towards the, what they argue, US military perceives as a possibility of achieving first strike capability with none or minimal possibility for taking damage from return fire. If you read the paper you will see that their argument is well supported.
While the authors themselves are writing from the point of view that this is a worrisome situation, I end up wondering, given what I wrote at the top of this post, whether this is not in fact a positive development. Despite all the faults of Western nations, I would feel far more comfortable in a world where Western civilization, Western values, and the Western way of life weren't threatened in the future on various fronts. But with a resurgent Russia carrying out technical upgrades to early warning systems and their nuclear arsenal, as well as a China with a very high economic growth, the window of opportunity for the US to take advantage of its nuclear primacy is short. As difficult as it is to say, I would rather millions die now than billions more suffer in the future.
Actually short for Prunesquallor, after Mervyn Peake's character in Gormenghast.
I also like how he hasn't published any peer reviewed paper since his PhD dissertation. Is that a scientist?
I remember during my undergrad in USF (the Tampa one, not SF) eating some popcorn while sitting on a bench in campus. Two squirrels came up to get some of the popcorn I had dropped on the ground. The problem is that they started fighting over it, surprisingly viciously, right around my legs. One managed to jump on my bare ankle and sink it's teeth--deep enough that I couldn't stop the bleeding and had to go to the campus clinic. The doctor said that in his 15 years working there, I'm the only one to have come after being attacked by squirrels.
I'd say a zoologist like Desmond Morris is a much more reliable reference on cats, and he is in agreement with the grandparent and disagreement with you.
- Disabled people or that need help in any way
Vote by mail.
- Those away on business
Vote by mail.
- Those living abroad
Vote by mail.
You miss the point. It shouldn't be about opinions, but verifiable facts. Credentials are useless in establishing truth.
who can't be bothered to go to a polling station to have an effect in an election?
I can only see the minor inconvenience of having to get off your ass in order to vote as a great way to filter out those that obviously don't care enough (and are likely as informed about the issues).
Automatic chemical foam extinguishers.
genuinely funny, even in the sad context of violence and oppression.
He didn't figure out arbitrary computations, just both addition and multiplication. That other computations can be built on top of those primitives has been known for much longer and is nothing new. So while in a sense your comment is formally correct, it is misleading.
Note that hardware generators are not guaranteed to produce great random numbers either, since they have non-uniform distributions. It's best to mix them up (in a feedback mode) with some other thing that has a good distribution, like a mersenne twister or coupled logistic functions.
Exactly. However, even here there is an implicit assumption--the one-time pad is truly random. If you use a crappy pseudorandom generator, your encryption will be vulnerable.
That is not correct. We're not simply dealing with chunks of encrypted data in a database here. We're talking about actual arithmetic operations. If your system supports at least addition and multiplication on encrypted data, then you can implement any other computation based on these.
Not with any proper encryption I can think of.
Then release binaries made with the Intel compiler. It's a better optimizer than MSVC (and gcc) whether on Windows or Linux.
One can always use the Intel compiler on Linux. The performance of compiled code is comparable to that the Intel compiler produces with its Windows version.