Slashdot Mirror

I think ORNL and PSC know a lot more about supercomputing than you (or Internet rag pundits) do. As others have noted, there are real reasons for Big Iron.

Clusters are great for certain problems but for heavy computation -- think simulating two galaxies colliding or earthquake modeling -- off the shelf clusters don't cut it.

They're not wasting tax-payer money unless you consider basic researcher a waste.

Goodbye Van Allan Belt, Hello Cosmic Rays . . .

Perhaps in the upcoming standarization they will finally switch to so called "jumbo frames", aka raise the maximum amount of data that can be sent in one chunk. As the singaling rate has gone up from 10Mb-1Gb, there has been a 100x increase in signaling rate and therefore a 100x decrease in the amount of time it takes one packet to cross the network. Since we are still using the same paltry sizes, cpu usage goes way up and throughput is somewhat capped. Switching to a larger frame size would allow higher throughput and lower CPU utilization. Many networking vendors have started adding support for larger frame sizes into their products for these reasons, but being added to the official standard would greatly increase the adoption of such jumbo frames.
For more info, see:
http://sd.wareonearth.com/~phil/jumbo.html
http://www.psc.edu/~mathis/MTU/
http://www.nwfusion.com/columnists/2004/0105tolly. htm

How about an anatomy lesson over the network? When people talk about the kind of applications that high-speed networks enable, tele-medicine usually pops up. You can't demo surgery but you can demo a virtual, real-time corpse!

The Visible Human Browser

I hope to work on things like Internet2, or in a large business environment after college.

Three words: Pittsburgh Supercomputing Center.

You might try looking for universities that operate a GigaPoP like the one at Pittsburgh Supercomputing Center (associated with CMU and Pitt). One CMU undergrad did an internship there (see the image and caption titled "Undergrad Excellence") and was hired after graduation.

You'll find other GigaPoPs listed on the Internet2 site.

You don't harvest extremophiles for these substances. You model the substance and find a synthetic analog

For you foreigners, this is a 'Pitt'.

It all boils down to this...
Pitt versus Penn State
Culture versus Agriculture

... and we have a supercomputers

Recent simulations suggest that planets form in hundreds instead of millions of years.

Does anyone know how this tech works? Won't it be limited by the electronics it's attached to?

Photonics promises to give us an all photon path but I don't think anyone's close to making an entire processor with photonic crystals yet.

Seti@Home already blows away all other supercomputers on the planet

Where did you get this idea? I'd honestly like to know. It's very misinformed.

There are many problems for which Seti@Home distributed-style computing would be worthless. One of the major selling points of real supercomputers are the interconnects between nodes. Big problems require fast interconnects so that nodes don't have to wait for data from other nodes working on other parts of the same problem. The Internet is way too slow for problems that have massive amounts of data that need to be passed between nodes -- such as modeling earthquakes or the earth's magnetic field.

A slighly more accesible artcile on aquaporins and ion channels:
Precious Bodily Fluids
about computational models.

bringing up the classic..
Is Hell Endothermic or Exothermic?

It goes head to head with Pitt's supercomputer here. Then it will be crushed and all the time/energy will be wasted.

This seems to be a good choice

From the above link: "This puzzle was inspired by a misleading statement attribued to Danica McKellar of The Wonder Years:" What's really amusing as that McKellar went on to get a BS in Pure Mathematics at UCLA, and write the foreward to "The Complete Idiot's Guide to Calculus."

Here's one for a triangle. No need for the sig thing, eh. Just doin' my job as a citizen of the world.

That being said, there are dozens of good reasons why manholes are round. The best one being their resemblance to donuts.

here

Another poster on this thread found the link, I had no idea what to search google for.

This particular answer always bothers me. Sure, it's simplistically true, but a whole family of shapes exists that has the same property but does not have the unfortunate property of spinning in place. For example, assume a vehicle stops on a manhole cover with a (powered) tire off-center on the cover. When the driver presses the throttle, the tire exerts a force on the manhole cover that gives it a tendency to rotate. Instant loss of traction.

Also, other shaped covers could posses a flange - the manhole would have a smaller maximum dimension than the flange, preventing the cover from falling down the hole. Squares or triangles would require unreasonably large flanges, but octagons wouldn't.

My guess is that a variety of factors (shape of manholes, ease of manufature, ability to roll the covers) lead to round manhole covers.

Does this research: New Light on Dark Matter, count as a prediction of these observations?

Simple. For many applications, OpenMosix isn't the solution. If you are looking for distributed workload (say, I need to run 15 serial jobs on a cluster shared with a bunch of people who need to do the same), OpenMosix works like a champ.

But if you need to do tightly-bound parallel computing, particularily with MPI (the article did mention MPICH after all), OpenMosix is pretty much not worth the effort. You can't do load balancing with MPI on OpenMosix because you can't move the jobs around - the sockets don't move right and MPI curls up and dies.

So, if you think clustering means "throw a bunch of machines out there and run a bunch of serial jobs on them", OpenMosix is great for clusters. If clustering makes you think of big, parallel machines, OpenMosix isn't the answer.

On a side note, some supercomputing center in Minnesota, or somewhere like that where its really cold in the winter, pipes out their heat into the parking garage to help the cars start. Also, the Pittsburg Supercomputing Center's heat output is equivalent to 169 pounds of coal an hour!.

Here's a direct link to PSC's article, which does -not- require registration (bah).

As mentioned by another post, we're talking about "Jupiter-like" gas giants, not Earths. The reason it can't take millions of years: "The problem with [the current model], however, is that if the formation process takes too long, nearby stars will, in effect, boil off the gas envelope."

This bit from the story:

Paramount's latest sci-fi thriller, The Core... depicts a world beset by just such a polar reversal, with radiation sweeping the planet.

There is solid evidence for gravity waves. Google for "binary pulsar gravity" if you want to find more about this piece of evidence.

Well, just a quick note, Pittsburgh is very much NOT a steel city any more. Its not quite the high-tech city that somewhere like Boston is, but still is pretty high-tech. (I'm orig from Pgh, but live in Boston now).

I don't even think that there are any working steel mills within 10-15 miles of the city, maybe even further. There is, however, CMU, the supercomputing center, UPMC medical center and a whole bunch of other research facilities.

For a comparion of magnetic pole shift vs other theories of polar and crusty disturbance, check out this page which picks apart the psuedo science of it all. There are a lot of wacko theories on what pole shift means, and a lot of it is based on lack of evidence and mis-conceptions.

It is intereting to note that, the earth's core is rotating faster than the surface crust to begin with.

There is this concern: The magnetic field acts as a shield against solar particles, etc. No field = no shield. Weak field = weak shield. This could be an issue with solar flares. Some folks are concerned that the field may be in the process of failing

Also, if the poles were to flip suddenly, many creatures that navigate magnetically could be affected. A full magnetic reversal could cause massive ecological problems across the whole of the Earth. If this took place slowly enough, each generation of creatures would learn to navigate with its' current situation.

The lameness filter won't let me post it here. I've put the necessary changes to your firewall script here. I used the same tool to generate my ipchains firewall and added this. I can play Quake3 and Wolf MP test just fine.

Here are some links to more information:

• List of the Top 500
  
• Official TCS1 Status Page(but not yet updated)
  
• TCS1 User Documentation: The Terascale Computing System
  
• Slashdot article announcing Birth of a Terascale Baby (Oct. 4, 2000)
  

Here are some links to more information:

• List of the Top 500
  
• Official TCS1 Status Page(but not yet updated)
  
• TCS1 User Documentation: The Terascale Computing System
  
• Slashdot article announcing Birth of a Terascale Baby (Oct. 4, 2000)
  

Notice also that IIRC the three sigma result was data from one detector.

I think it was almost 4 sigma from the Aleph detector that was driving the result. The NYT article is referring to data from the other detectors that has now been added in, driving the total significance down a bit to around 3 sigma.

Surely LHC will not be much cleaner than the experiment at Fermilab... doesn't H stand for Hadron? Anyway, computers get faster all the time which allows for much more sophisticated cuts and analysis to deal with the horrendous background.

Yes, H=Hadron (meaning proton). Fermilab is also a proton machine. The LHC will be the messiest machine ever built, worse than Fermilab due to the higher energy. Oh, actually, I think RHIC is probably the messiest machine ever built. Take a look at these event displays. One of them has almost 9000 tracks in one event! Looking at these makes me glad to be in theory!

--Bob

I don't think these people are wasting CPU time.

Doing this distributed over the Internet, however, is unlikely.

I don't think these people are wasting CPU time.

Doing this distributed over the Internet, however, is unlikely.

Is this a bad thing? I don't know. The Mellon Institute was involved in the development of many consumer products, including cornflakes and innerspring mattresses, as well as the GR-S synthetic rubber formula that helped win World War II.

This page on the PSC website gives detailed descriptions of the planned uses of this supercomputer...

Including Storm Prediction, Protein Folding, Turbulence Studies, Earthquake Preparedness, AIDS Research, Cardiac Fluid Modeling, Oceanic Phenomena, Electromagnetics and Fluid Dynamics.

They've also got some pretty neat animations of some of all of the above.

This page on the PSC website gives detailed descriptions of the planned uses of this supercomputer...

Including Storm Prediction, Protein Folding, Turbulence Studies, Earthquake Preparedness, AIDS Research, Cardiac Fluid Modeling, Oceanic Phenomena, Electromagnetics and Fluid Dynamics.

They've also got some pretty neat animations of some of all of the above.

According to their website... "PSC operates five supercomputing-class machines: a 512 processor Cray T3E, two eight-processor Cray J90s, a four-processor Alphaserver 8400 5/300 system, and an Intel cluster with 10 4-processor compute nodes."

This page provides a description of the work researchers plan to do with the new supercomputer.

The center is a joint venture between Carnegie Mellon, The University of Pittsurgh, and the old Westinghouse Electric company.

It's also intersting to note that the PSC & CMU formed the NCNE Gigapop that provides the internet to CMU, PITT, WVU, and Penn State.

The PSC has a release here

I was involved with the pittsburgh supercomputing center in high school. We were given a grant for processing time, something like $40,000, to compute the heat loss of my community due to improper insulation. Admittedly, I was on the fray of the group but I know they have been using massively parallel systems for a while. They also had an Internet connection which is where I first used Lynx.

At that time they had a T3D and a "DEC supercluster" which was IIRC 256 Digital Alpha computers. They had some other supercomputers but I can't remember what they were. The supercluster was later upgraded to 512 processors. It seems that this is the same thing, updated and built by Compaq (who bought Digital).

The PSC has a release here

I was involved with the pittsburgh supercomputing center in high school. We were given a grant for processing time, something like $40,000, to compute the heat loss of my community due to improper insulation. Admittedly, I was on the fray of the group but I know they have been using massively parallel systems for a while. They also had an Internet connection which is where I first used Lynx.

At that time they had a T3D and a "DEC supercluster" which was IIRC 256 Digital Alpha computers. They had some other supercomputers but I can't remember what they were. The supercluster was later upgraded to 512 processors. It seems that this is the same thing, updated and built by Compaq (who bought Digital).

For a nice article on simulations performed for the Relativistic Heavy Ion Collider:

A Taste of Quark Soup

BTW, this research was done a T3E (which uses Alphas).

Sean

For a nice article on simulations performed for the Relativistic Heavy Ion Collider:

A Taste of Quark Soup

BTW, this research was done a T3E (which uses Alphas).

Sean

Thought people might like to know this. Carnegie Mellon University has its own Linux distribution called Andrew Linux. We originally started out with RH 4.2 but have gradually made changes such that it no longer resembles it. The distro has provision for necessary CMU like Kerberos, AFS, etc built it.

Guess each college will make modifications to tailor Linux to its own needs. A common distro might just be unusable as each college has its own needs and probably a setup unique to itself.

The best names I ever saw for hosts was at the Pittsburgh Supercomputing Center, which sported two firewalls named Fear and Loathing.

In my office, we have named our systems after cartoon characters -- Warner Brothers only! :)

Porky - RH Linux (of course)
Marvin - Windows NT4 (world domination factor)
Taz - 450MHz Alpha
Droopy - old Sun SPARCstation 5
Speedy - Sun UltraSPARC
Wiley - G3/400 blue-n-white Mac

Considering that a 386 is far less powerful than a current top-of-the-line machine, but uses only a little bit less power, a point will quickly come where it's cheaper to buy a new machine than to pay for the electricity for a bunch of older machines.

Not to mention heat. At my former workplace, they got 10 rack-mounted quad-processor Xeons with the intention of clustering them. They were planning on testing the cluster in a small room in the basement before moving them to the main machine room. Plans had to be changed because of the amount of heat that 10 machines put out. The heat produced by 10000 machines would be truly phenominal.

TCP Tuning information for many operating systems (including Linux, BSD, and Microsoft's offerings) may be found at the Pittsburgh Supercomputing Center and NLANR Engineering Services