You have to remember that object orientation and such are all human creations. Processors don't think in objects, for that matter they don't really even think in functions. They think in memory locations, and jumps to those locations. Doing OO code means a whole messy layer the compiler has to go through to translate that in to something the processor actually understands.
That being said, Linux is written in a somewhat object-oriented style. Structs with pointers to functions and so on. It's just that some of the OO stuff is done manually by the programmer (setting up the pointers during init and destroy, calling functions via pointers instead of member functions etc.), and not automatically by the compiler.
I guess if one were to start a kernel project today, C++ would be a serious option. But when Linus got started in 1991 (?), there was no C++ standard, and g++ was crap.
As I understand it this sort of thing can be done on just about any kind of computer. And at every university I've ever been to there's usually stacks of old pcs laying around.
As opposed to running email and word, HPC is one of these things where CPU power actually matters. Those 500 MHz PC:s aren't worth the hassle to set up and maintain. Not to mention that heterogeneous hardware (which a random bunch of discarded PC:s probably is) is a nightmare to maintain and program efficiently in parallel.
Most clusters consist of quality rack servers from a reputable vendor. TCO matters, not the cost of the hw alone.
Not to many are using Fedora or Slackware on some white box with parts from Best Buy to do HPC. They have been altered to specifically run on hardware that was made specifically for this, and even then management of it is not exactly simple. Not that I believe that 2003 Server will suddenly change that but just using Linux somewhere does not automatically make it the cheapest way.
The "standard" cluster these days is standard rack servers from a reputable vendor, along with a Linux distro tailor-made for cluster usage such as Rocks or OSCAR. Typically the only nonstandard hw, if any, is a high-speed network (Infiniband, Quadrics etc.).
And I believe the correct answer to your question is Traditionally it has been done by tuned versions of commercial Unices which added to the base cost of the OS over and above the very expensive custom built hardware.
Perhaps in the mid-1990'ies yes..
Recently Linux has become able to do many of these tasks by similarly being modified at a significant cost running on the same expensive custom hardware.
No. With the exception of the high-speed network card I mentioned above, the rest of the hw and sw are bog-standard. Of course there are exceptions, e.g. SGI, Cray, NEC, IBM etc. but then we're talking "real supercomputers" and not commodity clusters (the market MS is aiming at).
The recent HPC installation using mostly off the shelf parts (they didn't use Ethernet) was the one at Virginia Tech and that ran OS X, not Linux.
Not to piss on OSX, but Mac clusters are probably outnumbered 100:1 by Linux clusters.
Another, perhaps even more popular Linux cluster distro is Rocks Clusters.
While I don't have personal experience with OSCAR, Rocks is really good. These days, doing a cluster with a "normal" distro is insane. I think MS will have to think long and hard before they come up with something equally easy to install and manage as Rocks.
That being said, I think MS is not targeting Win CCS at academic supercomputing, which has a long history of using Unix/Linux, but rather they want to expand HPC to business customers who otherwise have a 100 % MS environment.
Plutonium is much, much harder to detonate than uranium, which is what makes uranium enrichment so attractive to relatively low-tech bomb-makers.
OTOH, producing weapons grade plutonium is much simpler than producing weapons grade uranium, as you don't need isotope enrichment. Just a simple natural uranium fueled graphite reactor.
Couldn't this be worked around with virtualization? I.e. run both Vista and a free OS on the same box, communicate over TCP/IP. Kludgy, yes, but better than nothing I guess.
I wonder if anyone has tried that for real. Some sort of multiple server system up and running when someone puts a bullet through one without the system missing a beat. Now that's a video that would get some attention, both for the insanity and technical merit.
Funny you should say that. HP just did it with their high end storage array. See here.
Cyrus machines, which will share SAN storage via Lustre.
Have you tested that this provides good performance? AFAIK Lustre is designed and optimized for massively parallel applications doing sequential IO on huge files (e.g. HPC apps using the MPI-IO API). Sounds like maildir (lots of tiny files) is the exact opposite.
See e.g. article about a university system. Also Cyrus supports load balancing and failover via murder, although backends (the actual file stores) still remain single points of failure.
I'm in a similar situation. I had been using debian since 1997, but I got really annoyed at the ever lengthening release cycles, and I didn't want to use testing or unstable either due to periodic breakage.
Actually, my current Ubuntu system is a direct descendant of the original debian I installed in 1997. I have *never* reinstalled the system; tells something about the quality of debian and apt-get dist-upgrade! When getting new hardware it's just much simpler to cp -a the existing system, edit fstab and the grub config, than to reinstall.
Oh, and my parents are nowadays also happy Ubuntu users.;-)
I wonder whether the Cascade vector processor will really be a stand-alone vector processor or actually a co-processor?
Paradoxically, I believe that the major problem in making a really good vector processor is in pushing the envelope in single thread performance as well (Amdahls law and all that). In comparison, parallelism is easy. So in that sense, it would make sense for Cray to rely on AMD and x86 market volume to get good single thread performance very cheaply, and then concentrate resources on making a good vector co-processor. AMD licensing cache-coherent HyperTransport might fit in nicely here as well.
Also, from the software standpoint a vector co-processor makes more sense. Or else you would need an OS able to simultaneously run on different hardware architectures, or you need a vector chip capable of executing x86 code (uh oh)?
Don't hit yourself too hard about that though, since nobody really is a "real stock trader", as in consistently beating the market. Put your money in a index fund.
Certainly it's a laudable goal, but the cynic in me tells me that Sahlin, the politician, is just trying to score some greenie points. By 2020 she will be retired anyway, so it's not like she has anything at stake on this project actually amounting to anything.
Just like Sweden "decided" to phase out nuclear power in 1980 (hint: they still produce about 45 % of electricity by nuclear).;-)
You'll have to confirm this with the Folding@Home people, but the versions I saw previously were spitting out Gromacs http://www.gromacs.org/ messages, which is written in C++.
Nope. Gromacs is written in plain C, with alternative implementations for some inner loops in fortran and asm.
If you meant they can't go critical, that's wrong too - a reactor that can't go critical is also useless.
As a minor nitpick, that isn't necessarily true. Designs have been proposed where the reactor is subcritical at all times, and the reaction is sustained with the help of a particle accelerator. Suitable google-words might be "subcritical reactor transmutation spallation".
""" The waste material isn't actually that much of a problem. It's dangerous stuff, and you can't really "dispose" of it, I.E. leave it somewhere and forget about it. You've gotta live with it. Hundred of thousands of tonnes. But actually, it's not that much. Almost all of France's waste for the past 40 years sits in a place the size of a large warehouse. """
Well, the problem is that you have to store it for some 10,000 years. That's 2500 warehouses of pretty dangerous stuff, that you have to protect for a very long time.
Uh, how did one (1) warehouse suddenly become 2500 warehouses?
The "problem" didn't exist then (1970s and earlier)
No, since car safety standards were largely absent then, so any crash at higher than walking speed resulted in a crumpled piece of metal and corpses.;-)
The article is some mishmash of reality (wind power becoming competetive wrt fossil and the stirling solar systems are certainly interesting) and the most harebrained crackpot schemes around; Tom Bearden (Net loonie #1), "magnet power from vacuum", "blacklight power". Gee, it all sounds so credible.
My goals are pretty simple: I want to write applications that have a great look & feel that will primarily be pulling information from the web (think weather & news), play with that information and present it in interesting ways. I'd like those applications to be usable on the Linux and perhaps Mac OS X platforms.
In that case I'd recommend something like python combined with some gui toolkit such as wxpython or pygtk.
...into the guts of the machine
Since you're on some unix-like system, you could do worse than plain C and a few books (C:ARM5 by Harbison & Steel and Advanced Programming in the Unix environment by Stevens spring to mind). Some asm knowledge might be useful too.
As for tools, frameworks etc. there is of course an unending list of those. For an IDE, a like emacs code browser.
Nice to know your vest will stop a handgun but if a.223 can go right through it, it won't be to useful against a properly armed adversary.
Most modern armies use body armor and helmets even if they don't help against rifle bullets. Why? To protect against shrapnel (which iirc accounts for about 80 % of casualties in full scale warfare). So even if this doesn't protect against rifle bullets, it isn't exactly useless as long as it's an improvement over the standard kevlar stuff.
Slashdot Mirror
You have to remember that object orientation and such are all human creations. Processors don't think in objects, for that matter they don't really even think in functions. They think in memory locations, and jumps to those locations. Doing OO code means a whole messy layer the compiler has to go through to translate that in to something the processor actually understands.
That being said, Linux is written in a somewhat object-oriented style. Structs with pointers to functions and so on. It's just that some of the OO stuff is done manually by the programmer (setting up the pointers during init and destroy, calling functions via pointers instead of member functions etc.), and not automatically by the compiler.
I guess if one were to start a kernel project today, C++ would be a serious option. But when Linus got started in 1991 (?), there was no C++ standard, and g++ was crap.
Why is it insane?
Because you'll waste plenty of time solving problems that the Rocks team has already solved. See this presentation. Especially slides 38 and 47-57.
As I understand it this sort of thing can be done on just about any kind of computer. And at every university I've ever been to there's usually stacks of old pcs laying around.
As opposed to running email and word, HPC is one of these things where CPU power actually matters. Those 500 MHz PC:s aren't worth the hassle to set up and maintain. Not to mention that heterogeneous hardware (which a random bunch of discarded PC:s probably is) is a nightmare to maintain and program efficiently in parallel.
Most clusters consist of quality rack servers from a reputable vendor. TCO matters, not the cost of the hw alone.
Not to many are using Fedora or Slackware on some white box with parts from Best Buy to do HPC. They have been altered to specifically run on hardware that was made specifically for this, and even then management of it is not exactly simple. Not that I believe that 2003 Server will suddenly change that but just using Linux somewhere does not automatically make it the cheapest way.
The "standard" cluster these days is standard rack servers from a reputable vendor, along with a Linux distro tailor-made for cluster usage such as Rocks or OSCAR. Typically the only nonstandard hw, if any, is a high-speed network (Infiniband, Quadrics etc.).
And I believe the correct answer to your question is Traditionally it has been done by tuned versions of commercial Unices which added to the base cost of the OS over and above the very expensive custom built hardware.
Perhaps in the mid-1990'ies yes..
Recently Linux has become able to do many of these tasks by similarly being modified at a significant cost running on the same expensive custom hardware.
No. With the exception of the high-speed network card I mentioned above, the rest of the hw and sw are bog-standard. Of course there are exceptions, e.g. SGI, Cray, NEC, IBM etc. but then we're talking "real supercomputers" and not commodity clusters (the market MS is aiming at).
The recent HPC installation using mostly off the shelf parts (they didn't use Ethernet) was the one at Virginia Tech and that ran OS X, not Linux.
Not to piss on OSX, but Mac clusters are probably outnumbered 100:1 by Linux clusters.
Another, perhaps even more popular Linux cluster distro is Rocks Clusters.
While I don't have personal experience with OSCAR, Rocks is really good. These days, doing a cluster with a "normal" distro is insane. I think MS will have to think long and hard before they come up with something equally easy to install and manage as Rocks.
That being said, I think MS is not targeting Win CCS at academic supercomputing, which has a long history of using Unix/Linux, but rather they want to expand HPC to business customers who otherwise have a 100 % MS environment.
Don't forget the lesson of King Gillette.
When sales start to sag, add Yet Another Blade to the razor?
Plutonium is much, much harder to detonate than uranium, which is what makes uranium enrichment so attractive to relatively low-tech bomb-makers.
OTOH, producing weapons grade plutonium is much simpler than producing weapons grade uranium, as you don't need isotope enrichment. Just a simple natural uranium fueled graphite reactor.
Couldn't this be worked around with virtualization? I.e. run both Vista and a free OS on the same box, communicate over TCP/IP. Kludgy, yes, but better than nothing I guess.
I wonder if anyone has tried that for real. Some sort of multiple server system up and running when someone puts a bullet through one without the system missing a beat. Now that's a video that would get some attention, both for the insanity and technical merit.
Funny you should say that. HP just did it with their high end storage array. See here.
Cyrus machines, which will share SAN storage via Lustre.
Have you tested that this provides good performance? AFAIK Lustre is designed and optimized for massively parallel applications doing sequential IO on huge files (e.g. HPC apps using the MPI-IO API). Sounds like maildir (lots of tiny files) is the exact opposite.
See e.g. article about a university system. Also Cyrus supports load balancing and failover via murder, although backends (the actual file stores) still remain single points of failure.
I'm in a similar situation. I had been using debian since 1997, but I got really annoyed at the ever lengthening release cycles, and I didn't want to use testing or unstable either due to periodic breakage.
;-)
Actually, my current Ubuntu system is a direct descendant of the original debian I installed in 1997. I have *never* reinstalled the system; tells something about the quality of debian and apt-get dist-upgrade! When getting new hardware it's just much simpler to cp -a the existing system, edit fstab and the grub config, than to reinstall.
Oh, and my parents are nowadays also happy Ubuntu users.
I wonder whether the Cascade vector processor will really be a stand-alone vector processor or actually a co-processor?
Paradoxically, I believe that the major problem in making a really good vector processor is in pushing the envelope in single thread performance as well (Amdahls law and all that). In comparison, parallelism is easy. So in that sense, it would make sense for Cray to rely on AMD and x86 market volume to get good single thread performance very cheaply, and then concentrate resources on making a good vector co-processor. AMD licensing cache-coherent HyperTransport might fit in nicely here as well.
Also, from the software standpoint a vector co-processor makes more sense. Or else you would need an OS able to simultaneously run on different hardware architectures, or you need a vector chip capable of executing x86 code (uh oh)?
I am not (nor was ever) a real stock trader.
Don't hit yourself too hard about that though, since nobody really is a "real stock trader", as in consistently beating the market. Put your money in a index fund.
Certainly it's a laudable goal, but the cynic in me tells me that Sahlin, the politician, is just trying to score some greenie points. By 2020 she will be retired anyway, so it's not like she has anything at stake on this project actually amounting to anything.
;-)
Just like Sweden "decided" to phase out nuclear power in 1980 (hint: they still produce about 45 % of electricity by nuclear).
AMD exec says AMD is better than its competitor. Earth shattering news!
You'll have to confirm this with the Folding@Home people, but the versions I saw previously were spitting out Gromacs http://www.gromacs.org/ messages, which is written in C++.
Nope. Gromacs is written in plain C, with alternative implementations for some inner loops in fortran and asm.
If you meant they can't go critical, that's wrong too - a reactor that can't go critical is also useless.
As a minor nitpick, that isn't necessarily true. Designs have been proposed where the reactor is subcritical at all times, and the reaction is sustained with the help of a particle accelerator. Suitable google-words might be "subcritical reactor transmutation spallation".
"""
The waste material isn't actually that much of a problem. It's dangerous stuff, and you can't really "dispose" of it, I.E. leave it somewhere and forget about it. You've gotta live with it. Hundred of thousands of tonnes. But actually, it's not that much. Almost all of France's waste for the past 40 years sits in a place the size of a large warehouse.
"""
Well, the problem is that you have to store it for some 10,000 years. That's 2500 warehouses of pretty dangerous stuff, that you have to protect for a very long time.
Uh, how did one (1) warehouse suddenly become 2500 warehouses?
Will these ever see the light of the day?
The "problem" didn't exist then (1970s and earlier)
No, since car safety standards were largely absent then, so any crash at higher than walking speed resulted in a crumpled piece of metal and corpses.
The article is some mishmash of reality (wind power becoming competetive wrt fossil and the stirling solar systems are certainly interesting) and the most harebrained crackpot schemes around; Tom Bearden (Net loonie #1), "magnet power from vacuum", "blacklight power". Gee, it all sounds so credible.
My goals are pretty simple: I want to write applications that have a great look & feel that will primarily be pulling information from the web (think weather & news), play with that information and present it in interesting ways. I'd like those applications to be usable on the Linux and perhaps Mac OS X platforms.
In that case I'd recommend something like python combined with some gui toolkit such as wxpython or pygtk.
Since you're on some unix-like system, you could do worse than plain C and a few books (C:ARM5 by Harbison & Steel and Advanced Programming in the Unix environment by Stevens spring to mind). Some asm knowledge might be useful too.
As for tools, frameworks etc. there is of course an unending list of those. For an IDE, a like emacs code browser.
uh, you knew i was quoting Dirty Harry, right?
If you were, then you got it wrong. Dirty Harry used a
Nice to know your vest will stop a handgun but if a
Most modern armies use body armor and helmets even if they don't help against rifle bullets. Why? To protect against shrapnel (which iirc accounts for about 80 % of casualties in full scale warfare). So even if this doesn't protect against rifle bullets, it isn't exactly useless as long as it's an improvement over the standard kevlar stuff.