10 good Intel machines will not cost less than $10,000. For scientific work, I don't consider eMachines or your grey-boy solutions a "good" system.
So, I took the bait... I went to Compaq's site and spec'ed out an equivalent workstation. Note, I'm not souping up the video card or CD-ROM like the Apple workstations. No need to waste money.
...
I worked at Los Alamos this summer with 9 seperated machines and a 32 node cluster
None of the computers there were purchased from the point and click menu at Dell's web site.
So fuck off unless you have something useful to say.
The problem with this lies with Apple and distribution.
10 good Intel or AMD (dual procs) will cost less than $10,000. You don't need fancy video card or sound card, as the crap in the motherboard will do fine. You may not need a big hard drive depending on what you are doing, and you certainly don't need monitors, keyboards, mice, and dum dum dummmmm software licenses.
The macs will not get a 2:1 performance advantage in fact they will get no advantage, they will be at a disadvantage. For one thing, the bus speed is too slow.
Any way, for the same price you could get a 30x2 node P3 or XP cluster, and your jobs would get done in _at_least_ a fourth the amount of time.
Advocates of this approach claim that it has the potential to be faster than C++ and other native-code languages. A traditional optimizing compiler can only make decisions based on general predictions as to how the program will behave at run time. But if you watch the program's behavior, you have specific knowledge of what needs to be optimized.
That is what profilers are for. And if some Java VM could build a profiler into the interpreter/compiler then there is nothing to stop a profiler from being built into a C/C++ compiler.
I hate how posts like this get modded up on slashdot. It is in no way factual. Moderators are just being way to subjective. You can't vote something up because it appeals to you. We don't get to vote for the truth. This is the kind of stupid closed mindedness that... argh
Though there is a possibles side effect. You couldn't release nightly binaries, because the Intel compiler takes so long to compile optimized code and the Mozilla project is so large.
And the already small pool of UNIX Mozilla developers would get even smaller.
I know of several people switching to WindowsNT because
I know many people who have already switched from old school UNIX to Linux a long time ago. cough Fermi Lab cough Everyone I know in Los Alamos.
I know several people switching from Alphas to PPC/OSX but I sure as hell wouldn't extrapolate that out.
a) intel systems are extremly cheap (compared to architecures optimized for number crunching like RS6K)
Yes. That's exactly why everyone switched to Linux while you weren't looking.
b) compilers available for NT produced MUCH faster code, e.g. Digital fortran. (Yes, I know... but still a lot of excellent scientific computing software is written in fortran77, e.g. LAPACK)
...same with linux. I'll give you a hint, serious projects aren't compiled with g77.
When it comes to numerical simulation, run times in the order of weeks are not unusual, so a performance penalty of 50 percent is simply unacceptable.
Yes, performance is important. But stability is even more important. Programs that take weeks and months to run on a thousand node cluster just can't be run on an NT cluster. Do you think all the Linux clustering technology was invented to make the movie Titanic?
Smaller jobs can be done in NT, but it is difficult. You have to get everything running stable (which is possible, but it doesn't come out of the box that way) and running stuff remote stuff over multiple computers requires relearning and there aren't as many tools available.
Any way man, where have you been? Linux has taken over serious scientific computing. The price/performance ratio is just too good for P3 clusters running Linux.
I run at 1600x1200 with large, scalable truetype and type1 fonts and while without AA the fonts don't look bad, with AA the fonts appear to have perfectly, continuous, crisp edges and are not fuzzy at all.
Also, with the way XFree is set up now, you can have AA on/off for different sized fonts, so if you are at a low resolution you can turn AA off for tiny fonts and on for larger fonts.
And taking things to their logical extreme. The gravistar contains a universe that contains n gravitars that contain universes that contain f(n) gravitars each...
and turtles all the way down
And that would end up with exponentially infinite energy.
So then you would have to make absurd claims, like an infinite loop from universe to gravistar... and back to the begining
Maybe I am an idiot, but what equation is this. Is there a constant missing?, the left hand side is in newtons and the right hand side is in kg/(s^2 m).
This, coupled with the conservation of angular momentum, says that a true zero volume singularity can't exist through traditional mathematics, since as volume gets closer to zero everything to the right of the divisor grows. Now, whether you believe x/0 equals infinity or take the usual mathematical stance of saying that it simply can't be done it still equals a very unique situation. The mass at the center of a black hole must occupy some volume, after all, it's twisted enough spacetime in along with it to contain all the mass it holds.
he faster a black hole rotates the higher it's gravity, since the relative distance of the particles is so small and angular momentum refuses to decrease you have more relative velocity in a much smaller volume, with plenty of mass. Think of an electromagnet, lots of turns in a real real small space plus a ton of current will make one hell of a magnetic force, though confined to a very small area.
Why can't angular momentum be conserved
L = r x gamma p
then as r->0, gamma->Inf, because v->c
so a limit can exist
the same thing happens in particle physics by the way
Another thought for you, some black holes have jets of matter screaming out of their poles. When a gravstar spins, at the poles the matter doesn't have the same relative velocity as the matter at the equator, and when enough of a differential is created, poof, incoming matter is squirted upwards and outwards along the poles. Even if it's not matter escaping from the inside of the hole, the accretion disk would be forced much harder inwards perpendicular to the poles than drawn in along the axis of rotation, and the resulting pressure is god's own jet engine.
The jets of matter do not come out of the black hole its self, please be careful the way you state things. That is how misinformation spreads.
Re:You can still get sucked in
on
Black Holes Disputed
·
· Score: 2, Informative
Yes, he is sure. And if he is not, then I am.
That is the problem with black holes, they are singularities. Mathematical points, infinitely dense. It even says so in the article.
That is one of the problems with black holes, and why these guys are looking for something different.
Particle physicists have the same problems with electrons (they appear point like with no structure)
Feynman had the similar problems that he couldn't get rid of.
"if it is closed or open the equations lead quickly to massive expansion or sudden collapse."
On the universal time scale, what is 'sudden'
Last I heard, and I haven't payed attention in about 2 years, evidence was suggesting that Hubble's constant was getting larger over time... the universe is expanding faster and faster
Slashdot Mirror
the statement "a force of 1000G" is completely incoherent.
there are three 'g' variables in gravitational theory, not one of them is a force
'g' lower case, vector: is the acceleration due to gravity on the surface of the earth
'G' uppercase, scalar: is the universal gravitational constant that is the proportionality constant in Newtons famous equation
'G' uppercase, tensor: is the sometimes used notation for the Einstein curvature tensor
actually that would be a Thai accent my asian ignorant friend
So, I took the bait... I went to Compaq's site and spec'ed out an equivalent workstation. Note, I'm not souping up the video card or CD-ROM like the Apple workstations. No need to waste money.
...
I worked at Los Alamos this summer with 9 seperated machines and a 32 node cluster
None of the computers there were purchased from the point and click menu at Dell's web site.
So fuck off unless you have something useful to say.
The problem with this lies with Apple and distribution.
10 good Intel or AMD (dual procs) will cost less than $10,000. You don't need fancy video card or sound card, as the crap in the motherboard will do fine. You may not need a big hard drive depending on what you are doing, and you certainly don't need monitors, keyboards, mice, and dum dum dummmmm software licenses.
The macs will not get a 2:1 performance advantage in fact they will get no advantage, they will be at a disadvantage. For one thing, the bus speed is too slow.
Any way, for the same price you could get a 30x2 node P3 or XP cluster, and your jobs would get done in _at_least_ a fourth the amount of time.
That is what profilers are for. And if some Java VM could build a profiler into the interpreter/compiler then there is nothing to stop a profiler from being built into a C/C++ compiler.
2.0.XX is stuper stable
2.2.XX is BSD stable stable
2.4.XX is hardened WinNT stable
2.5.XX is Win95A stable
but seriously all applicable bug patches get backported and no new features are added so old stable releases are the really stable releases.
I hate how posts like this get modded up on slashdot. It is in no way factual. Moderators are just being way to subjective. You can't vote something up because it appeals to you. We don't get to vote for the truth. This is the kind of stupid closed mindedness that ... argh
proprietary != closed
nVidia != open
This is not a statement against Apple, but you are using the wrong words, or are making false claims
http://www.go-mono.com/faq.html#licensing
Though there is a possibles side effect. You couldn't release nightly binaries, because the Intel compiler takes so long to compile optimized code and the Mozilla project is so large.
And the already small pool of UNIX Mozilla developers would get even smaller.
I know many people who have already switched from old school UNIX to Linux a long time ago. cough Fermi Lab cough Everyone I know in Los Alamos.
I know several people switching from Alphas to PPC/OSX but I sure as hell wouldn't extrapolate that out.
a) intel systems are extremly cheap (compared to architecures optimized for number crunching like RS6K)
Yes. That's exactly why everyone switched to Linux while you weren't looking.
b) compilers available for NT produced MUCH faster code, e.g. Digital fortran. (Yes, I know ... but still a lot of excellent scientific computing software is written in fortran77, e.g. LAPACK)
When it comes to numerical simulation, run times in the order of weeks are not unusual, so a performance penalty of 50 percent is simply unacceptable.
Yes, performance is important. But stability is even more important. Programs that take weeks and months to run on a thousand node cluster just can't be run on an NT cluster. Do you think all the Linux clustering technology was invented to make the movie Titanic?
Smaller jobs can be done in NT, but it is difficult. You have to get everything running stable (which is possible, but it doesn't come out of the box that way) and running stuff remote stuff over multiple computers requires relearning and there aren't as many tools available.
Any way man, where have you been? Linux has taken over serious scientific computing. The price/performance ratio is just too good for P3 clusters running Linux.
What resolution are you running in <= 800x600 ?
I run at 1600x1200 with large, scalable truetype and type1 fonts and while without AA the fonts don't look bad, with AA the fonts appear to have perfectly, continuous, crisp edges and are not fuzzy at all.
Also, with the way XFree is set up now, you can have AA on/off for different sized fonts, so if you are at a low resolution you can turn AA off for tiny fonts and on for larger fonts.
I agree.
...
... and back to the begining
And taking things to their logical extreme. The gravistar contains a universe that contains n gravitars that contain universes that contain f(n) gravitars each
and turtles all the way down
And that would end up with exponentially infinite energy.
So then you would have to make absurd claims, like an infinite loop from universe to gravistar
and turtles all the way down
The nature of hollow spherical shells in classical physics relies heavily on Gauss's law, which in turn relies heavily on Euclidean space.
Am am not argueing with you, and you have stated that it is not exact, but I must say that I am much more hessitant. In fact I am highly skeptical.
Maybe I am an idiot, but what equation is this. Is there a constant missing?, the left hand side is in newtons and the right hand side is in kg/(s^2 m).
This, coupled with the conservation of angular momentum, says that a true zero volume singularity can't exist through traditional mathematics, since as volume gets closer to zero everything to the right of the divisor grows. Now, whether you believe x/0 equals infinity or take the usual mathematical stance of saying that it simply can't be done it still equals a very unique situation. The mass at the center of a black hole must occupy some volume, after all, it's twisted enough spacetime in along with it to contain all the mass it holds.
he faster a black hole rotates the higher it's gravity, since the relative distance of the particles is so small and angular momentum refuses to decrease you have more relative velocity in a much smaller volume, with plenty of mass. Think of an electromagnet, lots of turns in a real real small space plus a ton of current will make one hell of a magnetic force, though confined to a very small area.
Why can't angular momentum be conserved
L = r x gamma p
then as r->0, gamma->Inf, because v->c
so a limit can exist
the same thing happens in particle physics by the way
Another thought for you, some black holes have jets of matter screaming out of their poles. When a gravstar spins, at the poles the matter doesn't have the same relative velocity as the matter at the equator, and when enough of a differential is created, poof, incoming matter is squirted upwards and outwards along the poles. Even if it's not matter escaping from the inside of the hole, the accretion disk would be forced much harder inwards perpendicular to the poles than drawn in along the axis of rotation, and the resulting pressure is god's own jet engine.
The jets of matter do not come out of the black hole its self, please be careful the way you state things. That is how misinformation spreads.
Yes, he is sure. And if he is not, then I am.
That is the problem with black holes, they are singularities. Mathematical points, infinitely dense. It even says so in the article.
That is one of the problems with black holes, and why these guys are looking for something different.
Particle physicists have the same problems with electrons (they appear point like with no structure)
Feynman had the similar problems that he couldn't get rid of.
So, uhm, any number between zero and infinity?
But I want those goat sex pictures and links clearly seperated from all others
Maybe we need a new moderation
-10 Goat Sex
I program in C far more than LISP ... do not take me for an idiot. (not that returning a struct is a good solution)
Try this, have a function that takes an n-dimensional vector as input and an f(n)-dimensional vector as the output.
Yes, I could do this in C, but it would be a "trick".
functions that have multivariable output
in C I have to do tricks
in LISP it is natural
"Hmm... I think everything is energy."
You forget all the different kinds of charges.
You can't throw away 3/4 of the fundamental forces.
It was a rhetorical question.
I appologize for being a smart ass, I should have just said no.
"if it is closed or open the equations lead quickly to massive expansion or sudden collapse."
... the universe is expanding faster and faster
On the universal time scale, what is 'sudden'
Last I heard, and I haven't payed attention in about 2 years, evidence was suggesting that Hubble's constant was getting larger over time
snd3d.dll is from MSN messinger.
I just looked through the files in binary
The exe doesn't contain any ASCII text strings
looks like a VB program
Am I the only person here who reads word files with less?
garbage garbage garbage
text text text
garbage garbage garbage
What's cool about this is that all of the stuff that was deleted is still in there for me to read Mostly just bad sentences and stuff though.