Slashdot Mirror
All 500 of the World's Top 500 Supercomputers Are Running Linux (zdnet.com)
Freshly Exhumed shares a report from ZDnet: Linux rules supercomputing. This day has been coming since 1998, when Linux first appeared on the TOP500 Supercomputer list. Today, it finally happened: All 500 of the world's fastest supercomputers are running Linux. The last two non-Linux systems, a pair of Chinese IBM POWER computers running AIX, dropped off the November 2017 TOP500 Supercomputer list. When the first TOP500 supercomputer list was compiled in June 1993, Linux was barely more than a toy. It hadn't even adopted Tux as its mascot yet. It didn't take long for Linux to start its march on supercomputing.
From when it first appeared on the TOP500 in 1998, Linux was on its way to the top. Before Linux took the lead, Unix was supercomputing's top operating system. Since 2003, the TOP500 was on its way to Linux domination. By 2004, Linux had taken the lead for good. This happened for two reasons: First, since most of the world's top supercomputers are research machines built for specialized tasks, each machine is a standalone project with unique characteristics and optimization requirements. To save costs, no one wants to develop a custom operating system for each of these systems. With Linux, however, research teams can easily modify and optimize Linux's open-source code to their one-off designs. The semiannual TOP500 Supercomputer List was released yesterday. It also shows that China now claims 202 systems within the TOP500, while the United States claims 143 systems.
From when it first appeared on the TOP500 in 1998, Linux was on its way to the top. Before Linux took the lead, Unix was supercomputing's top operating system. Since 2003, the TOP500 was on its way to Linux domination. By 2004, Linux had taken the lead for good. This happened for two reasons: First, since most of the world's top supercomputers are research machines built for specialized tasks, each machine is a standalone project with unique characteristics and optimization requirements. To save costs, no one wants to develop a custom operating system for each of these systems. With Linux, however, research teams can easily modify and optimize Linux's open-source code to their one-off designs. The semiannual TOP500 Supercomputer List was released yesterday. It also shows that China now claims 202 systems within the TOP500, while the United States claims 143 systems.
Linux makes it to the desktop, of a supercomputer.
There is no second reason.
Linux is not used because it's better, it's used because it is cheaper.
In the end, cheaper almost always wins over better.
Unix never made inroads on the desktop.
This might actually be harmful if people think Linux is complicated or designed for heavy hardware they may not consider it for desktops and use cases involving desktop apps.
Linux has been ready for the desktop since about 1999, before that there were dependency issues and hardware wasn't always supported. Now hardware is more likely to be better supported on Linux than on Windows. I'm writing this on Windows but that's only because Windows came on this machine, I'll be installing Linux when I have a week of downtime.
Enlightenment is probably the best looking desktop software anywhere, it's customizability makes it hard to include with distros but it should be considered as evidence that it's not user-friendliness or beauty holding Linux back.
I think it's a bit sad to see Linux software becoming overly simplified in the wake of Apple's success the way other software is.
Linux needs to remain the enthusiast and expert operating system more than it needs broad acceptance. Look what happened with the internet, Linux is great without ads, malware and other problems I associate with popularity.
That said Linux skills are still hugely undervalued and not taught in schools which needs to change. A Linux machine is still your best bet that your machine will still be runnning with data and apps updated but not broken after 10-15 years.
Linux was originally made in Finland.
When all you have is a hammer, every problem starts to look like a thumb.
From what I know about the windows kernel it couldn't scale upwards well enough to run in this league. And If I remember correctly one of the key goals of Linux was to make sure it could scale well on big iron systems.
We still don't know if you can successfully beowolf cluster a bunch of the old Microsoft Barnies though.
I read at +2. If your post doesn't reach that level I will not see or respond to it.
oblig: other OS's are Finnished?
How far has the discussion quality fallen? Apparently this low, even without a political bent.
That time when Linus's wife could not be Rickrolled because her Linux box had no Flash capability was a searing tragedy in the annals of computer history.
I deny that I have not avoided attaining the opposite of that which I do not want.
So, the Top 500 list of computers was dominated by many Variants of Unix, with a little sprinkle of other weird stuff (among those, VMS). Which is not a monoculture
Then, as the other weird stuff waned, Windows took it's place (for a short while). Not directly as a replacement of course, but rather as a percentage of Top500 systems.
On the other side of the fence, Linux began to take increasign market share of the Top500 because of low cost, shallow learning curve from *nix, and posibility to modify source code, in an accelerated path to become a monoculture (at least where the Top500 is concerned).
And now, finally, we are on a monoculture in the Top500, with Linux all the way in the Top500... No *BSD, no AIX, HP-UX, or Solaris. Just Linux all the way.
Better not catch anyone complaining about Chrome Monoculture, Windows Monoculture, or Android monoculture! M'kay? ;-)
*** Suerte a todos y Feliz dia!
What'd you expect it to run? Windows?
This sig intentionally left blank.
five years ago, 3 of the top 500 did run windows, and in 2011 4 did.
Odd statement, considering Microsoft mantra declares Linux is far more expensive than Linux.
I think you got that backwards
I think he got it just right.
There was an "HPC" edition of windows 2003, and microsoft managed to sponsor a few places to build clusters using it that made it into the top500 list...
I don't recall anyone ever using it of their own volition tho, only if microsoft were paying, and at least one of those clusters was a dual boot experiment which climbed 50 places in the ranking when booted to linux.
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
I'm not even sure what you are asking here. Do you truly have no idea how a GPL works?
Anyway, you have this exactly backwards. The reason Linux became popular during the parallel supercomputing "revolution" (and I say this as a modest expert, at least at that time) is because it IS an open source operating system, so you could hack the kernel, write your own kernel drivers, fix things like networking bugs or system balance issues, and handle memory at a very primitive level. You got then, and can easily get now, the complete source of the OS and all of its device drivers, although the latter has been a constant source of contention between hardware mfrs who think that a device driver that makes their hardware run is some sort of "trade secret" and the keepers of the Linux kernel. Over decades (at this point) the mfrs have largely given up and actively help with kernel drivers instead of insisting on binary-only distributions. This played a critical role in the development of early parallel supercomputers once Linux had its first kernel capable of symmetric multiprocessing with two (and rapidly more) CPUs or (later) cores, or both. That would be roughly kernel 2.0, although there were still serious issues with race conditions, (network) driver interrupts and lockups, memory management, and so on, through 2.0.4+ -- really they went on forever as the 2.0 kernel wasn't truly symmetric, handled interrupt locking "badly", and took a lot of revision and some new paradigms to smooth out and stabilize. Ah, those were the days...
Microsoft, on the other hand, made you sign away your firstborn child in order to get a copy of the OS source -- even as a research institution. If (say) your network drivers were slow, or locked up while multiprocessing, you were SOL. You COULDN'T fix it. You couldn't even find the bug. And it wasn't worth the effort -- even if you sacrificed a goat and got the source -- to learn to work with the source because it changed at MS's whim and all your work could go down the tubes at any moment and if you DID develop anything that ran on their system in some "custom" fashion, you ran into serious issues if you wanted to share it. You COULDN'T share your work with anybody else, not unless they had a surplus of goats or firstborn children too.
"Anybody" (with a need and decent programming chops) could join the linux kernel list and communicate directly with the main kernel developers and report bugs, contribute fixes or drivers, etc. There was a lot of healthy debate about what needed to be fixed, or improved, first, second, third etc, as well as just how to go about fixing them -- sometimes it required substantial redesign and had to wait for a major bump (and a lot of testing). You could of course hack/fix your own kernels or add your own device drivers, or fix broken drivers, or mess with internal "tuning", and I and many others did, but behind the public scenes the actual kernel developers -- the heart of linux, as it were -- made steady, inexorable progress.
By the year 2000, Linux had made serious inroads into not only the top 500, but there were literally uncounted small clusters that weren't fast enough (or weren't architected correctly) to crack the top 500, which relied on things like the Linpack benchmark to determine who to include. There were lots of folks who didn't USE linear algebra in their computations who built massively parallel compute farms with many different architectures and purposes who didn't even have the benchmark software installed (or give a shit) about their "ranking". Both PVM and MPI were fully ported onto Linux and most of their ongoing development was taking place on Linux boxes. Additional tools for management and job distribution and much more were developed -- on mostly Linux boxes, but yeah, there were still SGIs and Sun Microsystems clusters and much more out there. They suffered -- badly suffered, terminally badly suffered in pretty much all cases -- from being much, much more expensive than over the counter Intel or AMD box
Even when the experts all agree, they may well be mistaken. --- Bertrand Russell.
The *BSDs will keep on doing what they have always done. Run well with minimal upkeep and not beta test features on production releases. Under Linux the mentality is if something compiles then ship it. I ran Linux in the 2.0.x kernel days. What they call Linux today is so far removed it might as well be a different operating system. Some distros don't even include tools like nslookup or traceroute anymore. Good luck installing that package if your default route isn't set. Oh and "route" has been changed to something else now for no good reason. What exactly was wrong with the old program and syntax?
Only the State obtains its revenue by coercion. - Murray Rothbard
I think the cliffhanger ending is the editors attempt to bring back slashdotting. There was once a time, when sites would be brought to their knees by a front page story on Slashdot. These days noone reads TFA, so the concept of slashdotting has faded from memory.
Actually he abandoned his homeland in search of warmth.
#DeleteChrome
... to fully appreciate all the features of the latest Enlightenment desktop.
#DeleteChrome
For most parallel problems, it's possible to divide them and send each piece to different computers, rather than a different core on the same computer. For even more highly parallel problems, using a GPUs to do the computation is even faster.
With 100 gig ethernet, we're starting to see networking speeds closer to bus speeds on motherboards themselves and it's cheaper, faster to scale (especially dynamically), and probably more fault tolerant (node fail? Send the job to a different node) to use more computer nodes rather than using more processors in a single computer.
Distributed computing has almost made supercomputers irrelevant -- except for people with a hole in their pocket. Folding@home is more powerful than anything on their list while we have no idea what monster of a compute clusters work inside Google or Facebook -- but given the open source software they have released (e.g. Facebook's 360 degree video stitcher) and how slow they are on a single machine -- the only way they'd be usable on their site is if you have a massive cluster.
-- Political fascism requires a Fuhrer.
Anybody still believe Linus Torvalds about how Linux was just for fun?
Of course. Linux was just for fun; now he makes a living out of it. A person's motivations for doing something don't have to remain exactly the same for the whole time they do it.
The technical term for this is "botnets".
Sent from my ASR33 using ASCII
shhh.. No one knows that all the TOP 500 supercomputers run on EMACS..
Bullshit.
Linux is used because it's far, far, FAR more flexible, less resource intensive and more efficient than Windows, while supporting and making good use of vastly larger amounts of RAM and CPUs.
If you baseline is one of the proprietary Unices, it's still more flexible, less archaic and more familiar to users while supporting a wider range of hardware while being infinitely cheaper.
After optimizing his computer to the point the CPU couldn't produce enough heat to warm his house anymore.
Trying to become famous by taking photos. Visit my homepage please.
I'm rather sure that means since the super computer builders are building their own OS out of Linux, they don't have to supply anybody the source as their not sharing it. However I may be mistaken...
You are mistaken. Top 500 shops are regular contributors to mainline Linux development, with test cases, patches and more than a few core developers. They do it because they benefit from it, and they save money that way, they don't need to carry patches. And they aren't "competing" in the commercial sense, they just want the best system they can have, and that means, play with the community.
anybody here that knows the contract of Linux that can verify this?
Contract??? You really don't get it, good luck with that.
When all you have is a hammer, every problem starts to look like a thumb.
Price is a positive thing of course but not why it is used - the cost for OS software in a supercomputer would be a fraction of the hardware and infrastructure costs anyway.
The thing is that Linux have excellent scalability when it comes to I/O throughput, this is something that many companies and individuals have worked hard to achieve. So it is possible to adapt an OS installation to be suitable for extreme throughput.
The compute nodes themselves doesn't really need a proper operating system (and many supercomputers/clusters in the past had extremely limited systems) and in fact any OS overhead is processing power wasted. Userspace programs using the MPI to communicate is the norm. And again the adaptability of Linux is an advantage, customizing a small efficient kernel with the necessities and nothing else is easy.
But the most important thing is that people already have done the adaptions and that those are available for others to use. Sure there are system specific things that have to be changed anyway but that would be the case for any large cluster machine. The closest to plug and play one can come.