Slashdot Mirror

200TB of Xserve RAID storage (link includes pictures)

Text of the article:

The University of Wisconsin - Madison has deployed 35 5.6TB Xserve RAID storage arrays in a single research installation as part of an ongoing scientific computing initiative.

The Grid Laboratory of Wisconsin (GLOW), a partnership between several research departments at the University of Wisconsin, have installed almost 200TB, or 200,000GB, of Xserve RAID arrays.

As a comparison, 200TB of storage is enough to hold 2.75 years of high definition video, 25,000 full length DVD movies, 323,000 CDs, 20 printed collections of the Library of Congress, or over 1000 Wikipedias.

The GLOW storage installation is physically split between the departments of Computer Sciences and High Energy Physics. Each Xserve RAID is attached to a dedicated Linux node running Fedora Core via an Apple Fibre Channel PCI-X Card and is either directly accessed via various mechanisms, such as over the network via gigabit ethernet, or aggregated using tools such as dCache.

The storage is primarily used to act as a holding area for large amounts of data from experiments such as the Compact Muon Solenoid (CMS) and ATLAS experiments at the Large Hadron Collider at CERN.

(hopefully not bad PR):

You did fine. ;)

CMS (= compact muon solenoid) is actually quite big detector. Its main purpose is to find the so called Higgs boson.

See, that's the sort of thing I was talking about. The CMS home page doesn't describe this at ALL. It has a FAQ page.. which promptly goes into details about the construction of the detector and how big it is without ever explaining why the thing is being built. A wikipedia link at the bottom eventually explains it all, but this is a rarity in my experience. It's written for physicists, by physicists.

Part of the problem I have as a non-physicist is that whenever I have to tell someone where I work, they immediatley want to know what the laboratory does, and why. It's difficult to explain the experiments when all you know is that they're building the biggest magnet ever.

Eventually my explanations fail to satisfy, and 9 times of 10 the conversation ends with someone asking "and my taxes are paying for that??" Public interest in theoretical research labs is already pretty damn low, and near as I can see a lack of explanation in layman's terms only hurts it further. Most folks are willing to accept that some types of study may never result in something they can buy at the store, but I also think they'd appreciate having a way to understand why it's important anyways.

Well, I am a physicist and here is some additional information (hopefully not bad PR):

LHC is the biggest and most powerful particle collider ever built. It is a proton-proton collider that collides proton beams together with 14 TeV (tera electronvolts) center-of-mass energy (if memory serves).

CMS (= compact muon solenoid) is actually quite big detector. Its main purpose is to find the so called Higgs boson. The existense of the Higgs boson is required by the Standard Model of particle physics (one good book on the basics of particle physics (for people who already understand quite a bit of physics and math) is: Francis Halzen, Alan D. Martin: Quarks and Leptons: An Introductory Course in Modern Particle Physics). CMS, as most other particle physics experiments has an onion-like structure. The innermost layer is called a tracker which is used to (surprise, surprise) find the tracks of the particles produced in the collision. There is also a magnetic field in the tracker so the curvature of the particle tracks can be used to determine their momenta. The next layers are called electromagnetic and hadronic calorimeters. These are used to measure the energies of the particles. And finally there are the muon chambers that are used to detect the muons (muon is like an electron but only heavier).

There are also other big detectors in the LHC experiment like e.g. ATLAS.

One good source of information on particle physics are CERN summer student lectures available in Real-media format.

Ah, the RCT for the CMS on the LHC in CH. Why didn't you just say that.

IANAPhysicist, but I work in proximity to them. So I know a little bit about this stuff.

RCT = A device that detects a particle after a collision happens in a particle accelerator, which "triggers" to the connected computer that something interesting happened.
CMS = Name of the experiment. Like NASA is the name of an organization.
LHC = A big particle collider being built at CERN, in Switzerland. Like Fermilab, but bigger.

Physicists are smart folk, but are hideous at PR. Most of the web pages intended to be a PR front fail miserably, and are indecipherable to anyone except physicists. There was even a movement a few years back to get physicsts to name their experiments in more public-friendly ways, which failed miserably.

Ah, the RCT for the CMS on the LHC in CH. Why didn't you just say that.

IANAPhysicist, but I work in proximity to them. So I know a little bit about this stuff.

RCT = A device that detects a particle after a collision happens in a particle accelerator, which "triggers" to the connected computer that something interesting happened.
CMS = Name of the experiment. Like NASA is the name of an organization.
LHC = A big particle collider being built at CERN, in Switzerland. Like Fermilab, but bigger.

Physicists are smart folk, but are hideous at PR. Most of the web pages intended to be a PR front fail miserably, and are indecipherable to anyone except physicists. There was even a movement a few years back to get physicsts to name their experiments in more public-friendly ways, which failed miserably.

altough it's an old news item and the blackhole is quit tiny. Keep in mind that this was only al "realy SMALL" particle collider. Currently there's an international project by europe china and america, the large particle collider, see http://lhc.web.cern.ch/lhc/ When ready (soon) this will be the biggest of it's kind and i wonder what things it might create. I don't like the idea of a blackhole near earth..

Have a look at ROOT. It is an object oriented data analysis framework with a C++ interpreter. It provides you with very powerful tools for doing all kinds of plotting (histograms and stuff), doing fits to data and storing data (so called ROOT files), etc... In addition to that it is free software (the latest version is licensed under the LGPL). It may not seem as easy to use as Matlab, but in the end I think ROOT is a lot more powerful.

On the other hand most secretaries and people like that would not find writing C++ scripts all that fun...

Heim has a somewhat unified theory about forces. Like Lorenz force, that is a force affecting charged particles, the Heim-Lorenz force affects any particle that has mass. (But the force still needs to be shown to exist in experiments)

Note that nobody's planning to make rockets out of this stuff any time soon...

[1] At least not if you're referring to stuff like antihydrogen. There are lots and lots of antiprotons produced, and they're "stored" by swinging them around and around the collider ring until they hit something (which is the reason they're produced in the first place).

[2] Of course, "in quantity" in this case means basically "more than one atom". I'm not sure just how many atoms they plan to store at a time, but it's certainly not nearly enough to do anything useful with (except possibly for spectroscopy). I mean, they purge the antihydrogen trap by turning off the fields and letting the antihydrogen annihilate against the trap walls. Poof! With little to no effect on the trap walls themselves (aside from probably accumulating the usual radiation damage that high-energy detectors like these have to put up with over time, of course).

Why is 15TB of storage news? (And they don't even the full 15TB yet!)

What about the 200TB of Xserve RAID storage for a single project at the University of Wisconsin, which has been up and running for over half a year?

And no, this isn't a project serving a whole campus or an entire university student body. This is one single research project operated by one entity. Oh well, I guess supporting the Large Hadron Collider isn't as cool as South Park. ;-)

There are some really nice APIs out there for computational work.

For example, those folks over at CERN don't seem to do any computational work, but when they have to, they like to use this:

Open Source Libraries for High Performance
Scientific and Technical Computing in Java

http://hoschek.home.cern.ch/hoschek/colt/

There are lots of other very interesting and handy APIs out there for computational work. I do bioinformatics, and can rely heavily on work other people have already done, and done well.

Perl is one of the last ones I'd foist on someone else who's not a programming professional.

Physicists are programming professionals. They deal with data sets, analysis problems, and hardware configurations that are way beyond the cutting edge. They build their own supercomputing clusters, write their own grid processing systems, build advanced data analysis frameworks, and fork their own Linux distros. At the physics lab where I worked for 15 years, if a physics grad student was incapable of learning a little Perl (and C, C++, Fortran, Java, TeX, and a couple of shells, and maybe some Python and Ruby) they didn't get their degree.

When launching a new physics project, it was a very serious concern which programming languages you chose to do your software development in. If you were conservative and went with a legacy language like Fortran because of all the pre-existing analysis software available, you had trouble attracting grad students to the project, because they wanted more marketable languages on their resumes. The reason is because if they decided to get out of physics one day, their strongest job prospects are in computing and data analysis.

Perl is one of the last ones I'd foist on someone else who's not a programming professional.

Physicists are programming professionals. They deal with data sets, analysis problems, and hardware configurations that are way beyond the cutting edge. They build their own supercomputing clusters, write their own grid processing systems, build advanced data analysis frameworks, and fork their own Linux distros. At the physics lab where I worked for 15 years, if a physics grad student was incapable of learning a little Perl (and C, C++, Fortran, Java, TeX, and a couple of shells, and maybe some Python and Ruby) they didn't get their degree.

When launching a new physics project, it was a very serious concern which programming languages you chose to do your software development in. If you were conservative and went with a legacy language like Fortran because of all the pre-existing analysis software available, you had trouble attracting grad students to the project, because they wanted more marketable languages on their resumes. The reason is because if they decided to get out of physics one day, their strongest job prospects are in computing and data analysis.

Wikipedia article on SCADA

See the SEAL Reflex project, the OpenC++ project, the PUMA project(now in AspectC++), Arne Adams' reflection library, and XVF by Kurt Stephens. These all work and provide introspection to some degree. There are other projects like Stroustrup and dos Reis's The Pivot and Vandevoorde's Metacode that may make it into future C++ standards to make it easier to provide good introspection support.

SEAL Reflex http://seal-reflex.web.cern.ch/seal-reflex/

OpenC++ http://opencxx.sourceforge.net/

AspectC++ http://www.aspectc.org/

Reflection library http://www.arneadams.com/index.html

XVF http://kurtstephens.com/research/paper/xvf_paper/

The Pivot talks http://charm.cs.uiuc.edu/patHPC/slides/stroustrup- a.pdf
http://www-unix.mcs.anl.gov/workshops/DSLOpt/Talks /DosReis.pdf

Metacode talk http://www.open-std.org/JTC1/SC22/WG21/docs/papers /2003/n1471.pdf

More seriously though, I heard someone describe the threat of an object accelerated to nearly the speed of light, and aimed at the earth. The resulting explosion would tear a hole right through the planet and make life unlivable.

Ummm, no. Humans currently do this all the time - it's called a particle accelerator. This is an ordinary, everyday thing at CERN, the RHIC at Brookhaven National Laboratory, and lots of other places.

LHC@home has not finished: it has only finished its current group of work units. From its October 28 news announcement: "The current group of studies is coming to an end. ... There will probably be a pause while the results are studied." This is similar to its August 24 news announcement: "The studies are almost done. ... We will then do post analysis to determine where there is more work needed." LHC@home uses the results of one set of work units to help it create the next set of work units. As the project's 4th status report shows, the project will have more work for a long time.

This feature of the project shows one of the strengths of BOINC: instead of LHC@home participants having to shut down an LHC@home-only computing client, start up a different project-specific client, and check the project's website regularly to see when new work units are available, they don't have to do anything: their BOINC client will automatically work on other BOINC projects until LHC@home has work units available, and will then automatically start working on LHC@home again.

Boinc itself isn't really a replacement for seti though, it is simply a manager
You choose which projects you wish to subscribe to, and how long you want any particular project to hog resources for and away you go.
At first i ran seti alone, but recently I have been running the Einstein@home and LHC@Home client on a 33% resource share basis with Seti.
Einstein, looks for spinning Pulsars and the LHC is a client from CERN running simulations of particles spinning around the new Six Track large hadron colider.
The LHC project has just finished sadly, but I think I'll move onto the Rosetta project, which is looking to work out various protein structures and interactions and how they can be used.

If, like me, you always fancied running a few other projects other than Seti but didnt want the hassle of manually deciding which client ot run then Boinc is a real boon and well worth the few minutes needed to set it up.

Have a go, I think you will like it!

The WWW per se was 'inveneted' at CERN by Tim Berners-Lee. Details here.

There was at least one that I am aware of. It was comparison of 3 DBs with respect to its application for quite a big project.
The report (made few years ago) is here:
http://dcdbappl1.cern.ch:8080/dcdb/archive/ttraczy k/db_compare/db_compare.pdf
However the machine seems to be down right now, so google html version:
http://64.233.183.104/search?q=cache:ijRVyqqJ_8EJ: dcdbappl1.cern.ch:8080/dcdb/archive/ttraczyk/db_co mpare/db_compare.pdf+dcdb+postgresql+mysql&hl=en&c lient=firefox-a

FWIW: The internet was never invented by Americans. The internet was actually invented by CERN, a Swiss multi-national european research group. They have moved off this and are now actually making antimatter. No, really, they are. See: http://public.web.cern.ch/Public/Welcome.html

You might find the following link useful.

CERN on WWW

>> develop a alternative to HTTP
HTTP came out of CERN which is an acronym for "Conseil Européen pour la Recherche Nucléaire"; Horror! European and it has a French name.
It started in 1954, in large part as a means of healing the wounds of the war. Since then much fine research has been done there, not least because of the free and open collaboration between the physicists of Europe and the rest of the world - including the US. That group of people will show huge competitiveness - my experiment is ALWAYS the most important - but none of this mean-minded "I did it, it's my ball" BS.
The US physicists I have worked with - from SLAC, Fermi, Brookhaven - have never shown parochial jealousies. Perhaps because they get out and see what the rest of the world is like. Mostly because that's what they do - try to find out what the world is like.

It was Tim Berners-Lee at CERN.

"...U.S. control, which stems from the country's role in creating the Internet as a Pentagon project and funding much of its early development."

The internet was invented by a CERN scientist in Europe. Visit

http://public.web.cern.ch/Public/Content/Chapters/ Spotlight/SpotlightAandD-en.html

for more. At the same time the contribution of the US may well justify their control.

By the way, read the 'Da Vinci Code' and the 'Angels & Demons' by Dan Brown!

Matyas

That's because developers don't want to pay for RHEL in order to build packages against it. Really, can you blame them?

No. I can't.

But you paid RedHat all that money. Tell them to get off their asses and package the software you want to use for that distribution you paid too much for.

I didn't pay anything and I am really glad I didn't. I am using Scientific Linux 4.0 (a distro built using RHEL4 srpms). I installed it because I had some compiler problems with Debian Sarge and Ubuntu. Their gcc version didn't compile Geant4 simulation toolkit properly and I didn't have time to fix things so I had to install something else and I just happened to choose Scientific Linux 4 (I figured that at least SL4 would have a compiler version that would work with Geant4). Now I have simulations and data analysis code on that system and it would be just too much trouble to format and install Debian, Geant4, ROOT data analysis framework and all the custom C++ scripts I have written during last few months. I won't switch to another distro before I have finished the project.

That's because developers don't want to pay for RHEL in order to build packages against it. Really, can you blame them?

No. I can't.

But you paid RedHat all that money. Tell them to get off their asses and package the software you want to use for that distribution you paid too much for.

I didn't pay anything and I am really glad I didn't. I am using Scientific Linux 4.0 (a distro built using RHEL4 srpms). I installed it because I had some compiler problems with Debian Sarge and Ubuntu. Their gcc version didn't compile Geant4 simulation toolkit properly and I didn't have time to fix things so I had to install something else and I just happened to choose Scientific Linux 4 (I figured that at least SL4 would have a compiler version that would work with Geant4). Now I have simulations and data analysis code on that system and it would be just too much trouble to format and install Debian, Geant4, ROOT data analysis framework and all the custom C++ scripts I have written during last few months. I won't switch to another distro before I have finished the project.

I also run the LHC@home project, as well as Einstein@home and Seti@home on both my machines (P4 1.8 and AMD64 3500+). Seems to work nicely.

http://lhcathome.cern.ch/

How about CERN? I got my first internship ("summer student programme") in 1997 and it was just great! Lots of fancy hardware, including the 7000-ton, 40-meter long and 20-meter high ATLAS detector, which sports the technologies to the limits, the 27-kilometer long LHC accelerator, buried 100 m below the surface, which will be filled completely with super-liquid helium (at 1.9K), and - last but not least - the place where GRID is being worked on. Have a look: http://humanresources.web.cern.ch/HumanResources/e xternal/recruitment/students/students.asp

Seriously, mainframes are so cool. And they offer patent protection, too.

Of course, it's needed some tuning so it wouldn' think that things that should be talking to multiple systems in a short time window don't get blocked...

WWW: later addon from MIT

Sorry to burst your bubble but WWW is a CERN invention (international organization part in Switzerland, and part in France). Check here and here.

Picture a world...Talking about "pictures" and the beginning of it all, here's the first picture ever to be clicked on in a web browser...

"How about Cern and Tim Berners-Lee? The initial Netscape release was basically the same as NCSA Mosaic which came before it."

Just to clarify:

CERN is the European Organization for Nuclear Research (the acroynm isn't English-language). Tim Berners-Lee "created" the original web browser, WorldWideWeb, while he was working there.

Mosaic was developed at NCSA, the National Center for Supercomputing Applications at the University of Illinois in Urbana-Champaign.

Marc Andreesen and Eric Bina were the original creators of Mosaic while they were students working at NCSA. Andreesen later founded Netscape Communications (originally Mosaic Communications) to try and build a company around the success of Mosaic.

You may have a look at this:
http://public.web.cern.ch/public/Content/Chapters/ AboutCERN/Achievements/WorldWideWeb/WWW-en.html
among others, includes the link to the proposal of the WWW made at CERN by Tim in 1989:
http://www.w3.org/History/1989/proposal.html
and refined by Robert Cailliau in 1990:
http://www.w3.org/Proposal.html

BTW, noone seems to remember about Robert Cailliau, the co-author of the thing...

Once again, the WWW was invented in EUROPE at CERN (the worlds LARGEST particle physics laboratory). The Internet without the WWW is about as useless as tits on a bull, get you facts right!

I've heard from older physicists that in those early years the scientist-computer match was quite popular.
(It still is, but, well...)

In the High Energy Physics (particle physics) world, people use ROOT or its predecessor PAW. ROOT's not GPL but a proprietary free license. ROOT is all C++; it's totally object oriented with a bewildering feature list. You can make really pretty graphs with it, and it exports to postscript among other formats.

ROOT and VTK

http://indico.cern.ch/ published under GPL

But of course, the Web came out of CERN.

That question is non-sense. You confuse programming languages with compiler implementation techniques.

A language is a syntax (how do I write it?) and a semantic (what does it do when I run it?). Languages never say anything about you are supposed to implement it, and so asking if a language is interpreted or compiled is just non-sense.

There are many implementations of the Java language, with many different execution strategies.

• Sun's implementation of Java is a Just-In-Time compiler (JIT).
  
• 
  SableVM is also an implementation of the Java language. It's execution strategy is to interpret the bytecode.
  
• 
  GCJ is another implementation of the Java language. It's execution strategy is to compile ahead of time.
  
• 
  DrJava is yet another implementation of the Java language. It's execution strategy is to compile the Java source code to Scheme code, then macro-expand the Scheme code to primitive Scheme, then compile that to bytecode, then interpret the bytecode.
  

• 
  The Ocaml language comes standard with three implementation: an interpreter, a bytecode compiler/bytecode interpreter pair, and an aggresively optimizing native code compiler. Somebody else also implemented a JIT.
  
• 
  The standard Python implementation is an interpreter. Starkiller is a Python compiler, and Psyco is a Python JIT.
  
• Last but not least, GCC is a C compiler (of course), and Cint is a C interpreter.
  

Now go in peace, and never say "compiled language" again.

That question is non-sense. You confuse programming languages with compiler implementation techniques.

A language is a syntax (how do I write it?) and a semantic (what does it do when I run it?). Languages never say anything about you are supposed to implement it, and so asking if a language is interpreted or compiled is just non-sense.

There are many implementations of the Java language, with many different execution strategies.

• Sun's implementation of Java is a Just-In-Time compiler (JIT).
• SableVM is also an implementation of the Java language. It's execution strategy is to interpret the bytecode.
• GCJ is another implementation of the Java language. It's execution strategy is to compile ahead of time.
• DrJava is yet another implementation of the Java language. It's execution strategy is to compile the Java source code to Scheme code, then macro-expand the Scheme code to primitive Scheme, then compile that to bytecode, then interpret the bytecode.

• The Ocaml language comes standard with three implementation: an interpreter, a bytecode compiler/bytecode interpreter pair, and an aggresively optimizing native code compiler. Somebody else also implemented a JIT.
• The standard Python implementation is an interpreter. Starkiller is a Python compiler, and Psyco is a Python JIT.
• Last but not least, GCC is a C compiler (of course), and Cint is a C interpreter.

Now go in peace, and never say "compiled language" again.

It was 3 or 4 years ago when I saw a 600 terabytes (0.6 petabytes) tape-based storage system at CERN.

PCI-X: a full-width (64 bit), full-speed (133MHz, quad-clocked) PCI-X bus gives 1064Mbyte/s peak bandwidth, shared between all devices on that PCI-X segment

Multiple PCI-X segments: this is already happening on some high-end boards (e.g. the Newisys board used in Sun's quad-Opteron v40z).

Network drivers which support NAPI (e.g. Intel's PRO/1000 NICs) are much more likely to achieve Gbit speeds in practice

Besides, other posters mentioned many other options (unattainable at the moment, though) which may offer a way out of the FTL problems. Many of modern scientific and technological problems are impossible to solve if you just try the most straightforward approach; the word "workaround" means exactly that :-)

The University of Wisconsin has deployed 200 TB of storage for support of similar types of experiments as part of the Grid Laboratory of Wisconsin.

Brief article, with pictures:

University of Wisconsin deploys nearly 200TB of Xserve RAID storage (Google cache)

The storage is used for, among other things, particle physics simulations in support of research projects at sites such as the Large Hadron Collider at CERN. More information on GLOW and its initiatives can be found here.

Text of the above article:

The University of Wisconsin - Madison has deployed 35 5.6TB Xserve RAID storage arrays in a single research installation as part of an ongoing scientific computing initiative.

The Grid Laboratory of Wisconsin (GLOW), a partnership between several research departments at the University of Wisconsin, have installed almost 200TB, or 200,000GB, of Xserve RAID arrays. As a comparison, 200TB of storage is enough to hold 2.75 years of high definition video, 25,000 full length DVD movies, 323,000 CDs, 20 printed collections of the Library of Congress, or over 1000 Wikipedias.

The GLOW storage installation is physically split between the departments of Computer Sciences and High Energy Physics. Each Xserve RAID is attached to a dedicated Linux node running Fedora Core 3 via an Apple Fibre Channel PCI-X Card and is either directly accessed via various mechanisms, such as over the network via gigabit ethernet, or aggregated using tools such as dCache.

The storage is primarily used to act as a holding area for large amounts of data from experiments such as the Compact Muon Solenoid (CMS) and ATLAS experiments at the Large Hadron Collider at CERN.

Aside from the GLOW initiative, the university also has Xserve RAID storage systems in use in other areas as well.

Full disclosure: I am the administrator of alienraid.org and am affiliated with the University of Wisconsin.

The University of Wisconsin has deployed 200 TB of storage for support of similar types of experiments as part of the Grid Laboratory of Wisconsin.

Brief article, with pictures:

University of Wisconsin deploys nearly 200TB of Xserve RAID storage (Google cache)

The storage is used for, among other things, particle physics simulations in support of research projects at sites such as the Large Hadron Collider at CERN. More information on GLOW and its initiatives can be found here.

Text of the above article:

The University of Wisconsin - Madison has deployed 35 5.6TB Xserve RAID storage arrays in a single research installation as part of an ongoing scientific computing initiative.

The Grid Laboratory of Wisconsin (GLOW), a partnership between several research departments at the University of Wisconsin, have installed almost 200TB, or 200,000GB, of Xserve RAID arrays. As a comparison, 200TB of storage is enough to hold 2.75 years of high definition video, 25,000 full length DVD movies, 323,000 CDs, 20 printed collections of the Library of Congress, or over 1000 Wikipedias.

The GLOW storage installation is physically split between the departments of Computer Sciences and High Energy Physics. Each Xserve RAID is attached to a dedicated Linux node running Fedora Core 3 via an Apple Fibre Channel PCI-X Card and is either directly accessed via various mechanisms, such as over the network via gigabit ethernet, or aggregated using tools such as dCache.

The storage is primarily used to act as a holding area for large amounts of data from experiments such as the Compact Muon Solenoid (CMS) and ATLAS experiments at the Large Hadron Collider at CERN.

Aside from the GLOW initiative, the university also has Xserve RAID storage systems in use in other areas as well.

Full disclosure: I am the administrator of alienraid.org and am affiliated with the University of Wisconsin.

[rant]If you "generally consider the internet as USA only", that probably says a whole lot more about you than it does about the internet. Are you aware you wouldn't even be reading /. if it weren't for the Swiss CERN, creating the WWW? Do you know that broadband penetration is as high (if not higher) in many EU countries as it is in the US? Don't you think it's about time for many Americans to drop the conceited attitude, and to look around and notice they're not alone on the planet?[/rant]

That being said, according to TFA, The origin of the zombie machines may change on a daily basis as machines can be infected anywhere in the world. CipherTrust has found that during April and May, the largest percent of zombie originations have alternated between China and the United States. In addition, during the first three weeks of May, approximately 26% of daily new zombies originated in the European Union, so let's not jump to any conclusions about Europe's supposed backwardness here. The figures may very likely show an entirely different picture again tomorrow, as they apparentely did just a few weeks ago.

...for the PowerPC eieio instruction.