Slashdot Mirror

Ah, but it is efficient. Planes are more efficient than cars because they can run on set schedules and carry more people per unit fuel. Trains are the same way, especially high-speed ones. With the right technology, trains beat planes for cost/mile on short-haul intercity passenger transportation, which is one reason why the Boston-Washington train route is profitable while others aren't.

Increasing speeds will help to improve the cost performance of trains. However, we should skip steel rails and go straight to InducTrak, a revolutionary maglev system with tracks that can be laid like steel because they need no electrification.

There was an experiment where a scientist used LOX and charcoal to see how fast it would burn - it esentially flashed in less than a second.

Experiment? More like "let's see how fast we can light a barbecue grill!" ;)

p

Yucca mountain is perfectly safe, but it's kind of stupid. We have the technology to reprocess most of the stuff into more fuel for reactors. Other stuff is useful in medicine and micro power sources. The remainder can be made safe via a process known as "Photoremediation".

The reason why this isn't done (save for some allowance for the second case I listed), is that the government considers it a threat to national security. Their problem with these options is that evil terrorists may intercept nuclear materials shipments, then use them for evil deeds. So their solution is to pile it all in a big cave somewhere. *sigh* Things are pretty bad when our own government doesn't understand.

Well, Hydrogen in gaseous and liquid forms already is transparent... but there's another form of hydrogen that is silvery and opaque: Metallic Hydrogen which can exist under extreme pressures such as the core of Jupiter and Saturn. It is theorized that metallic hydrogen would look a lot like molten lithium, perhaps even a lighter silver, almost white, if you could view it.

Huge misconceptions seem to abound here. FIRE does not represent the whole of US fusion research. There are dozens of other projects and laboratories around the country, most in academia and the national labs.

$2M/year is just for this ONE project.

The summary is extremely poorly written, and apparently the submitter thinks that the US is "canceling" all of its fusion programs, when in reality, ONE project of many is being canceled. The whole reason FIRE came about is because the US pulled out of ITER. Now we're back in, and FIRE could serve as a backup project potentially, but ITER is the focus in this particular line of research. But there are still many, many federally funded fusion research programs, projects, and laboratories around the US! We've spent $5 billion on projects like the National Ignition Facility (NIF) alone (only to be crucified by the Left...I guess you can't win).

Jeez. Wake the fuck up, or at least learn something.

While it is sad to see this program closed, the USA has at least 1 other fusion project active.

NIF, National Ignition Facility
http://www.llnl.gov/nif/

Check it out!

Does anyone know if the National Ignition Facility was part of the "FIRE" program? According to a previous slashdot article (referenced link on story now dead), they had high hopes of firing several Terawatt/TeraJoule lasers to produce fusion by 2010.

... he's just playing, because he's looking at the data visualization window, and not reading the /. page over to the right of the display:

http://www.llnl.gov/icc/sdd/img/images/DMX-Chromiu m-VisIt.jpg

This is a common problem on supercomputers: you have lots of users that want to run lots of jobs that have conflicting requirements for resources, and typically some dependancies between jobs and the like. Take a look at some of the scheduling and resource management tools available for supercomputers and maybe one of those will scratch your itch.

A couple pointers to get you started:

• SLURM, which while designed for Linux clusters is a good system and at least should seed a Google search (disclaimer: I work for LLNL and am on the user end of slurm, and I'm only speaking for myself here).
• Condor is a lot more than scheduling, but it does that as well.

Those are the ones I think it would be useful to look at for now. Most of the other systems are vendor specific.

-"Zow"

I am not sure on the details as I have not done this myself.

But in your situation I would be creating a make file to schedule the jobs with. Make can handle concurrency and with available patches can be made to distribute jobs to multiple nodes. Parallel Make Patches for GNU Make.

In a method like this I would recomend a small shared file system so that as you complete each job you can touch a file. This would allow you to continue from the point you left off, or if you wish, clear it out and start over.

btw, that SL8500 has what appears to be a max capacity of 90 Petabytes (!!!) so i'm wondering .. who would have that much data to backup?

What would have happened, in the heat of the moment, if instead of crashing a plane into the Pentagon, they detonated a small, dirty nuke.

Absolutely nothing. Dirty bombs are primarily scare tactics. They're actual ability as a tactical weapon has been highly overrated. Here's a good write-up for you.

that's because the other tools are PBS, Condor and LoadLever, all complete crap. Grid Engine is also crap, but at least doesn't have the 80s baggage.

Personally, I found PBS to be the best open source solution last time I had to choose, but that was just prior to the Sun buyout of GRD, so things may have changed. [My current employer rolls their own batch scheduler, so I haven't had a need to survey the field for a few years.] There are also some things Condor rocks at (cycle scavanging, userspace checkpoint/restart/migration) which none of the others even attempt, so it's definitely worth a look for some sites.

If your paying $$ for your batch scheduler, LSF pretty much trumps all of them, but the price is too steep for me.

I'll have to agree with you whole-heartedly. A Google search for "mesh compression" turns up just a few of the resources available on this kind of research. To me, this is like reading a Slashdot article about a new improvement to Windows BMP files called "JPEG compression."

The ROAM algorithm, for example, is one of a set of similar algorithms. It's specifically for terrain-type models and turns up in games now and then. (AFAIR, TreadMarks was the first big game to use it.)

Yeah. We have a relative who's always forwarding internet chain mail/urban myth letters.

I finally got pissed off and told her, that I wouldn't mind if she directed it to me specifically and asked, "Is this for real?", but if I got them as part of a chain, my wife and I would be very pissed off at her.

It seems to be working... I've gotten three "is this real?" emails in the past two weeks. I let her know if it's bogus or true, often with a Snopes reference (warning -- popups), or a reference to CIAC.

Sometimes just telling someone to check their facts first works (yeah, I know, about the weather report in Hell).

BTW, a few links for you:

http://www.spacedaily.com/news/nuclear-blackmarket -02d.html
http://www.llnl.gov/csts/publications/sutcliffe/

The end result is very few (if any) people would die from the radiological effects. Of course, maybe the public would know better if Nader had done something useful and taken the nuclear challenge.

Quote from the site:

"Diffractive telescopes using Fresnel lenses fabricated on thin membranes offer several advantages over telescopes using mirrors; thin membrane lenses are lightweight, packageable and space deployable. Transmissive diffractive lenses are much less sensitive to surface deformations compared to mirrors, and the chromatic effects due to the diffractive primary can be completely compensated for."

So, there's a use for 'em - cheap, large-scale space telescope objective lenses which are relatively robust compared to mirrors - no "Hubble trouble" with these. And if manufactured in segments, as LLNL is doing, it's easy to create *very* big lenses which fold up into a small space.. take a look at the demo lens at the bottom of the page.. wonder how many pennies THAT thing'd melt. ;)

Actually, I see a lot of people disclaiming the idea of using lenses for telescopes (too heavy, fresnels don't focus well, mirrors are cheaper and lighter) - however, this page (at Lawrence Livermore National Laboratory) describes a concept for a space telescope using a *five-meter* segmented fresnel lens as the objective, two spacecraft separated by a few km, and a smaller (movable) lens as an 'eyepiece', in effect creating a huge refracting telescope. Quote from the site: "Diffractive telescopes using Fresnel lenses fabricated on thin membranes offer several advantages over telescopes using mirrors; thin membrane lenses are lightweight, packageable and space deployable. Transmissive diffractive lenses are much less sensitive to surface deformations compared to mirrors, and the chromatic effects due to the diffractive primary can be completely compensated for." So, there's a use for 'em - cheap, large-scale space telescope objective lenses which are relatively robust compared to mirrors - no "Hubble trouble" with these. And if manufactured in segments, as LLNL is doing, it's easy to create *very* big lenses which fold up into a small space..

Thanks to your tax dollars, something similar is being used at the National Ignition Facility which is a HUGE facility dedicated to starting a fusion reaction by throwing as much energy into a central chamber as possible (all using lasers of course)

From what I see, they are not using fresnel lenses, but, rather, some sort of synthetic-diamond prism.

Remember that fresnel lenses aren't all they're hyped up to be. They are large, thin, and cheap. Those are the advantages which come primarily at the expense of optical quality. A conventional lens this size would need to be several feet thick, effectively weighing a couple thousand pounds.

Also, imagine polishing such a beast into the proper parabolic shape (not semicircular). It would take several years and millions of dollars. Most telescopes (the hubble included) don't have lenses that thick (alright.. . the hubble has a mirror - convex/concave lenses and mirrors are quite similar optically)

But you must not forget that plutonium is a heavy metal and toxic. Guess why we are proud to finally produce lead-free batteries, microchips and such? So disposal and processing of plutonium based products will be some nasty job. The radiation -- as you said -- is not that big of a problem, though inhalation of plutonium particles in almost the tiniest quantity will very likely lead to cancer, since plutonium is known to highly carcinogenic.
This and other articles state that the toxicity of Pu is not that critical, since it will be immobilized in sediments or soil. But I do not fully trust this argument. Pu just like Pb is stored in living organism to some extent, and will probably accumulate in animals along the food chain. Maybe this is not a problem in the short term, like it was a problem with lead from car fuels, but probably in the mid to long term.

Intel provides excellent Linux support for Itanium. Also if you use the Intel compiler, which Lawrence Livermore does, you get considerable speed boost on Intel CPUs.

See: http://www.llnl.gov/linux/linux_basics.html#compil ers

Intel can afford to provide little niceties like this. Can AMD? I doubt it.

Okay... imagine a Beowulf cluster of these! Actually, here it is.

LLNL built a supercomputer, and it's going to do things besides simulate nuclear weapons?

Quick, someone ring Satan and ask how the sno-cones are.

Don't be silly, Apple is a very small fish in a very big pond compared to real supercomputer companies like SGI.

What the big companies do is let you login to their machines free of charge, over the internet, and take their machines for a spin. See e.g. http://www.testdrive.compaq.com

On top of that, what the companies with a *serious interest* in high performance computing (like SGI and Cray) do is let you email your program to them: they will spend days or weeks (up to you) tuning it, telling you how it works on their hardware, and even telling you that you're better off with the competition, if that ends up being the case. What these companies understand very well, that Apple doesn't, is that people who spend $10M+ on computing equipment, typically with public money, need to show due diligence in their choice of hardware. This means THOROUGH evaluation: benchmarks, benchmarks, benchmarks - and not _standard_ benchmarks, but the benchmarks that matter: the user's own software.

Take a look here to see a real-life snapshot of this kind of process.

Any company that tried to "fudge the numbers" would be caught out, and that looks VERY bad. So the companies instead do all they can to help with the evaluation process, and hope that they get chosen. If not, there's always the next sale. A big supercomputer is sold somewhere every day or two, after all.

Virginia Tech's "X" doesn't come into this category: for example, no actual scientific work has been performed on the machine so far apart from benchmarks/system development, etc. Dr Varadarajan is probably going to get a pretty nasty grilling in a couple of years, when the University asks "so, what did we get for our money?" Some P.R., and not a whole lot more it would seem. I wish them luck...

Apple is not really a supercomputer company: and really, who cares? Apple has been, is, and will probably remain a great computer company for a good long while yet. But don't expect Apple to help you out a great deal if you pop them an email going "say, how does the G5 run on the ASCI Purple benchmarks?". That's just not what Apple are about. (example: they are happy to sell you a box with a 64-bit CPU, but they could care less that they don't have a 64-bit OS to go with it..)

I don't think the fat tree died with Thinking Machines. For example, MCR at LLNL uses a Quadrics fat tree. I imagine many sizeable clusters (way more than 64 nodes) use one. There's one link here, and the MCR link here but you can probably google for quadrics and fat tree to find some more. I'd be surprised if fat trees didn't show up in Myrinet / other interconnects, but you typically need to have a sizeable cluster before there's any point in calling it a fat tree.

(Oh, and if you meant something else entirely by fat tree, I apologize. I'm not too familiar with the particulars of the CM5 fat tree, so the Quadrics one is the only usage I'm aware of.)

50 trillion of calculations per second. Is that a synonym of flop (floating-point operation)? ...

How does this computer compares with the BlueGene/L (131,072 cpus, 0.5 Petaflops -estimated)? Don't be mislead by the name (*Gene)... this will be a computer for classified simulations (it will have a 1-2 year long "science run", for testing purposes with non-classified simulations).

Cheers...

Intercepting the OpenGL calls from non-modified OpenGL applications and rendering them to walls and CAVEs is a nice trick.

If this interests you at all, you also need to check out the open source Chromium project, which can do that, and much, much more. While it doesn't have the event tracking that VRiser appears to have, it has the ability to render to tiled displays, stereo displays, CAVES, do distributed sort-last compositing, OpenGL stream modification on the fly, parallel OpenGL submission, and a heck of a lot more. It supports high-speed cluster interconnects such as Myrinet, Quadrics, and Infiniband. It's also pretty easy to add your own OpenGL modification if you want to do something special.

As an example, check out this project that uses Chromium to split up live Quake games into an external isometric view.

(Disclaimer: I'm one of the Chromium developers, and my Lab helps pay the external developers to write this open source tool.)

Even in the server market, cutting on power consumption is getting more and more important. If you have a park of 1000+ machines in a data center, power consumption matters.

Altough not Pentium related, just imagine the efforts needed to power and cool down the BlueGene/L at L.Livermore. They have to cool down 131,072 CPUs! IIRC, BlueGene will use 1.5 MW, and should work at 0.5 Petaflops (as one guy from LLNL told me).

The article says that they used ttcp which is a memory-to-memory bandwidth testing program. Most would consider that unrepresentative of reality. On the other hand, today's supercomputers have a tremendous amount of memory (1.2 TB, 6 TB, 10 TB, 33 TB, etc.) so memory to memory is possible.

Others have suggested that disk speeds cannot sustain that rate. However, supercomputer disk arrays can easily keep up (4 GB/s or 32 Gb/s).

Finally, it is possible to achieve nearly the same result (multiple streams instead of a single stream) transfering real data (23.23 Gb/s).

[Bias alert: I am a member of the team that set a previous Internet2 Land Speed Record, Guinness World Record and won the "Bandwidth Lust: Distributed Particle Physics Analysis Using Ultra-High Speed TCP on The Grid" or "Moore's law move over" award at SC2003.]

Now, before you complain that the technology is not available to "mere mortals," let me point out that we first started experimenting with 1 Gb/s Ethernet at work 5 years ago. Now it is readily available at reasonable prices for consumer desktop machines. (Apple has had it standard in G4 desktops for 4 years.) The problem is not with consumer hardware, it is having access to true broadband (not cable modem or DSL), at least in the USA. Although your LAN may support 1 Gb/s, your download speed is limited to 1-3 Mb/s (cable) or 256 -786 Kb/s (DSL). (Your upload speeds are significantly lower.) Since the link provider has very little incentive to upgrade service, I doubt that will change very quickly.

So, yes it is possible. No you can't have it (yet)!

Why bother with something as dangerous as fluorine gas? Simple ultraviolet radiation should be sufficient.

Because you don't want UV rays, you want overkill. To demonstrate my point, allow me to quote a bit of a page from the Lawrence Livermore National Laboratory:

Direct skin exposure to pure fluorine can cause severe burns in 0.2 sec, and an exposure for as long as 0.6 sec can result in thermal flash burns comparable to those produced by an oxyacetylene flame.

After reading that, it is now clear that Fluorine is the right hazardous material for the job.

Is this similar or the same as this FreeBSD vulnerability? That one was fixed in 1998.

doesn't this CTO of cray remind you of someone?
"There IS no Linux in high-performance clusters."

"There IS no Americans in Iraq."

OMG! It's the former Iraqi mis-Informed-ation minister!

Especially when 2004 has been dubbed the year of the penguin, it's wreckless to claim that Linux can't be used in HPC's.
Hell, just look at the current top500 list. There's no Cray in the top 10 but there are two Linux based clusters there (and one based on OSX [FreeBSB based]).

Here's a few:
NCSA's IA32 Linux cluster
NCSA's IA32 Linux cluster
Space Simulator Clust at Los Alamos (SS51G based; makes me proud as I have a SS51G too)
Beowulf - used in many Linux clustering projects
Linux clusters at Los Alamos (they seem to have more than one)
Virginia Tech's Supercomputer X

As mentioned in the article, these ultra-pure glasses are needed for industrial lasers. Impurities or contamination in a laser glass can absorb the laser energy, be ioned, and explode within the laser glass. Tradionally, large scale glass melting is performed in ceramic refractory lined tanks. The problem is that glass melts are highly corrosive and a significant amount (any amount is significant when dealing with high powered lasers) of contamination will occur.

For the laser glass blanks used for the National Ignition Facility NIF Website, U.S. taxpayers supplied platinum lined tanks to prevent refractory contamination in the glass blanks. However, even platinum is soluble in the phosphate based glasses used for these blanks. Platinum colloids would result in catastrophic failure of the glass once the lasers where powered up So, a great deal of research has been performed by the glass science community (thank you DOE) to learn how to alter the glass chemistry via composition or processing parameters in order to incorporate these tiny levels of platinum into the glass structure and render it harmless in the finished product.

If significant quantities of these types of glasses could be made using containerless levitation, the expense of platinum lined glass tanks and challenges of neutralizing what little platinum contaminates the glasses would be eliminated.

No, it's because this is just about the only way to study actual fusion (not just computer simulations) under bomb-like conditions without using a bomb.

But you don't have to take my word for it. Here's what the LLNL itself says about their NIF:

NIF is crucial to the Stockpile Stewardship Program because it is the only facility that can create the conditions of extreme temperature and pressure - conditions that exist only in stars or in exploding nuclear weapons - that are relevant to understanding the operation of our modern nuclear weapons. In addition, NIF is the only facility that can create fusion ignition and thermonuclear burn in the laboratory. Nuclear fusion is the process that our modern nuclear weapons use to achieve their immense explosive power. The understanding of these conditions and the data provided by NIF will allow our nuclear stewards to assess and certify the aging stockpile without actual nuclear tests using supercomputer modeling tools.

Judging from the Windows XP backgrounds on these computers lets just hope they don't get the blue screen of death when they test the final setup.

http://www.llnl.gov/nif/construction/photoshow.htm l

National Ignition Facility home page
National Ignition Facility project status and photo gallery with lots of pictures
LLNL Science on High Energy Lasers

National Ignition Facility home page
National Ignition Facility project status and photo gallery with lots of pictures
LLNL Science on High Energy Lasers

National Ignition Facility home page
National Ignition Facility project status and photo gallery with lots of pictures
LLNL Science on High Energy Lasers

National Ignition Facility home page
National Ignition Facility project status and photo gallery with lots of pictures
LLNL Science on High Energy Lasers

I cannot help but think that it would be unwise to tick off the LLNL.

I cannot help but think that it would be unwise to tick off the LLNL.

OSCAR vs. other cluster software: HA-OSCAR is a logical development of other open-source cluster software out there. For instance, see SLURM, a package for scheduling jobs on a Linux cluster.

As a student at a major Big Ten University (tm) I can easily tell that your perception is a bit skewed. The old cliche "you get what you put into it" applies to many things in life, and computer science is no different.

My school's core computer science curriculum is in Java. Language of instruction is a moot point to a rather great extent. You can learn as much from a data structures class taugh in Java as you can from one taught in $language_of_choice. The idea is to learn how things work fundamentally, and then apply those ideas practically. A linked list in Java works the same as a linked list in C. Its not about Java being the "industry standard" as you call it, its about Java being a perfectly modern and capable programming language. The idea

Your next analogy of the cable repairmen almost prompted me to moderate your post as +1 Funny, but when I found out you were not joking I decided to write this reply instead. To even equate a cable repair person with a computer scientist is pure madness. Even if they were programmers, how is getting the cable modem working a good metric of "computer stuff in general" being "a lot less like a science or craft and more like a factory job", or even relevant to the discussion of computer programmers vs. computer scientists at all?

None of your points even remotely explain what you consider the fundamental problem: "why software sucks...why the programming "trade" sucks...why companies can send the jobs abroad to work for peanuts" The fact is not all software sucks, many people love their jobs in the industry, and these people are getting paid well to do their jobs. Most of the computer scientists you speak of don't work in the private sector, you can find them at government research institutions.

To say that these type of people don't currently exist, and that current CS curriculums can't produce scientists of this caliber is nothing short of ignorant.

I would not call that disinformation at all. I assume you are attempting to make a subtle distinction between management of the labs, and what one typically means by "belong to", not that you are attempting to troll or anything.

Prime Contract W-7405-ENG-48 (Contract 48 or Prime Contract) is executed between the University of California (UC or University) and the U.S. Department of Energy (DOE), and it is under this contract that LANL and LLNL function.

In other words, UC manages the labs under contract from DOE.

An outright banning cell phone from the workplace is extreme and should be met with a quick move to another company.

I'll have to mention that to my employer, who bans cell phones (and any personal electronics) in areas where classified information is discussed and processed.

An outright banning cell phone from the workplace is extreme and should be met with a quick move to another company.

I'll have to mention that to my employer, who bans cell phones (and any personal electronics) in areas where classified information is discussed and processed.

I've thought that this kind of phenomena is really electronic rather than nuclear.

Kind of like the tribo-electricity and tribo-luminescence. Purely electron transfer, some localized temporary ionization, but I would really doubt one could get as much eV as you typically need for nuclear reactions.

I could be wrong - perhaps the spherical symmetry of the collapse helps to localize the energy enough.

OTOH, maintaining spherical symmetry under some of these conditions can be difficult. Just ask the NIF folks if maintaining spherical symmetry is a cake walk.

Sonoluminescence: an Introduction

Single Bubble Sonoluminescence HOWTO

They aren't competitors. Notes is a collaboration/groupware suite.

And we aren't collaborating in a group right now? People don't use Intranets and Internet email for what they would have bought Notes for in the mid-90s? I knwo for a fact that that's what happens at non-Microsoft shops. w.g. Oracle doesn't use Notes internally. It uses Internet email, web-based solutions and some collaborative addons of their own.

They aren't competitors. XML is just one of many protocols that can be used to implement CORBA. Corba is an Architecture, XML is a data transmission format.

I was talking about XML in the large: XML+SOAP+WSDL, etc. Obviously these are both pitched as enterprise integration technologies and XML-based ones have a lot more traction in business today (think .NET and Axis) than CORBA does.

You don't (if you are sane) use a scripting language to write enterprise-level apps like finance or CRM software, or secure distributed systems, or high-performance numerical software.

GNU Enterprise is finance software written in Python. Secure distribute systems in Python? How about mojo nation or ZEO, or the MEMS Exchange or BitTorrent. High performance numerical software? You'd better tell someone down at Lawrence Livermore National Labs that they are insane because they show up at every Python conference and by now have spent millions on Python code. I don't see Java or C# mentioned on their list of key languages. Java in particular is a horrible language for that sort of thing. Do a Google for "Java Floating Point".

Look: you can understimate Python just as the Unix vendors understimated Linux. In the long run it doesn't really hurt anyone, even you. It is always more comfortable to presume that things will stay in the mental boxes we've built for them in our minds.