Slashdot Mirror

← Back to Stories (view on slashdot.org)

World's Most Powerful Private Supercomputer Will Hunt Oil and Gas

Posted by samzenpus on from the black-gold dept.

Nerval's Lobster writes "French oil conglomerate Total has inaugurated the world's ninth-most-powerful supercomputer, Panega. Its purpose: seek out new reservoirs of oil and gas. The supercomputer's total output is 2.3 petaflops, which should place it about ninth on today's TOP500 list, last updated in November. The announcement came as Dell and others prepare to inaugurate a new supercomputer, Stampede, in Texas on March 27. What's noteworthy about Pangea, however, is that it will be the most powerful supercomputer owned and used by private industry; the vast majority of such systems are in use by government agencies and academic institutions. Right now, the most powerful private supercomputer for commercial use is the Hermit supercomputer in Stuttgart; ranked 27th in the world, the 831.4 Tflop machine is a public-private partnership between the University of Stuttgart and hww GmbH. Panega, which will cost 60 million Euro ($77.8 million) over four years, will assist decision-making in the exploration of complex geological areas and to increase the efficiency of hydrocarbon production in compliance with the safety standards and with respect for the environment, Total said. Pangea will be will be stored at Total's research center in the southwestern French city of Pau."

26 of 135 comments (clear)

  1. 2.3 gigaflops? by shoppa · · Score: 3, Insightful

    2.3 gigaflops is on most everyone's desktop today. Maybe you mean 2.3 teraflops?

  2. Article and Summary have wrong units by fgodfrey · · Score: 4, Insightful

    This has to be 2.3 *peta* FLOPS not giga FLOPS. For instance, in 2010, an Intel desktop processor could do 109 gigaFLOPS (reference: http://en.wikipedia.org/wiki/FLOPS).

    --
    Go Badgers! -- #include "std/disclaimer.h"
  3. Oil by Anonymous Coward · · Score: 2, Insightful

    So.... why are we wasting the most powerful computer on non renewable sources of energy?
    Hunting for big oil just seems... wrong. Why cant we use this computer to find cures, track the stars, simulate atoms?

    1. Re:Oil by K.+S.+Kyosuke · · Score: 2

      Why cant we use this computer to find cures, track the stars, simulate atoms?

      Why, you'd like to try to split the computer?

      --
      Ezekiel 23:20
    2. Re:Oil by Overzeetop · · Score: 2

      Because oil is where the money is. And as it gets more scarce, it will be even moreso.

      Ever wonder why nobody seems to fight for ending hunger, or world peace, or finding alien life? Because I know you can't buy a yacht and a mansion on every continent by keeping people from starving to death.

      --
      Is it just my observation, or are there way too many stupid people in the world?
    3. Re:Oil by the+eric+conspiracy · · Score: 3, Insightful

      That's because there is plenty of food. The problem is political, not technical or economic.

    4. Re:Oil by manicb · · Score: 2

      Hi. I use supercomputers to model materials for next-generation solar cells. If you're working on semiconductors, you need to probe the electronic structure of crystals with junctions and defects; basic atomistic calculations aren't going to be enough (although they have their own uses and are definitely part of the picture). These calculations help us to work out why some promising materials don't perform well in experiments, and figure out which reactions will be suitable for cheap large-scale production and processing. Even simple atomistic models require a lot of power if your system is large enough and/or you want to look at complex processes like melting.

  4. Re:2.3 gigaflops? by shoppa · · Score: 2

    Crap, I got it wrong too. Not 2.3 teraflops either. 2300 teraflops = 2.3 petaflops.

  5. Re:umm? by K.+S.+Kyosuke · · Score: 3, Insightful

    When I saw "Panega", my first thought was "For a geological supercomputer, that's a nice pun on Pangea, isn't it?" Well, then I noticed that it's not a pun, just a typo. ;-)

    --
    Ezekiel 23:20
  6. This is how it starts by Anonymous Coward · · Score: 2, Informative

    I bet cold-pressed humans are a wonderful source of hydrocarbons.

  7. Pangea vs. Stampede by Anonymous Coward · · Score: 2, Informative

    Quite an impressive system in general terms, 2.3PF without accelerators says a lot about the size of this machine (48 racks):

    "Pangea is manufactured by SGI, built on the ICE-X platform. In a video, Total said that each blade contains four Xeon processors (most likely the E5-2600, which SGI uses), each with 32 cores and 128 Gbytes of RAM. Each M-Rack contains 72 blades, for a total of 288 processors, 2304 cores, and 9 TB of RAM. An M-Cell contains four M-Racks and 2 C-Racks for 288 blades, 1,152 processors, 9,216 cores, and 32 TB of RAM. In all, 12 M-Cells are used, with 110,592 cores, 442 TB of RAM, and 120 km of fiber-optic cable connecting it all up. Pangea also includes 12 bays, with 600 1-TB drives each, and 4 petabytes of magnetic tape for archiving data."

    A system of this size with accelerators would exceed easily 10PF, although I am not sure whether the particular workload to be ran on this beast would be suitable for any kind of accelerators (anybody has an idea on that?). Now I have a question: what is TACC going to do with so many Xeon Phi accelerators not delivering the promised performance? Will intel provide them with the second generation of MICs for free or will that upgrade cost another big chunk of taxpayers money?

    X.

  8. Re:Amateurs by interval1066 · · Score: 3, Insightful

    I couldn't care less about that nonsense. I'm waiting for the day they use this system to hunt humans. Dissenters, etc...

    --
    Python: 'And then suddenly you have a language which says "we're all stuck with whatever the whiniest coder wants".'
  9. Waste of computer power by Anonymous Coward · · Score: 3, Insightful

    Oil and gas?! How about Solar? Wind? GeoThermal? LFTR/Thorium? Why are we blowing computer power on dying industries!?

    What is this?! The Freakin' Flintstones?!

    1. Re:Waste of computer power by shadowrat · · Score: 3, Insightful

      Oil and gas?! How about Solar? Wind? GeoThermal? LFTR/Thorium? Why are we blowing computer power on dying industries!?

      Well, it seems like a waste of computing power to use it to find those things. I mean, we already know where the sun is. We already know where it's windy.

    2. Re:Waste of computer power by drinkypoo · · Score: 2

      Well, it seems like a waste of computing power to use it to find those things. I mean, we already know where the sun is. We already know where it's windy.

      Actually, we could know this much better, but it's remarkably hard to figure out without dropping a kajillion weather stations.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    3. Re:Waste of computer power by geekmux · · Score: 2

      Oil and gas?! How about Solar? Wind? GeoThermal? LFTR/Thorium? Why are we blowing computer power on dying industries!?

      Well, it seems like a waste of computing power to use it to find those things. I mean, we already know where the sun is. We already know where it's windy.

      Waste of computing power? Most likely.

      Waste of money to those invested in it to help feed a greedy and corrupt oil industry? Not hardly.

      Just follow the money. Oddly enough, it always seems to lead to the true answers.

  10. Re:Most powerful commercial supercomputer by h4rr4r · · Score: 2

    How about useful commercial work?

    How about just less single purpose?

    It is not like you can take a bitcoin rig and do anything worthwhile with it.

  11. So how does it work? by jones_supa · · Score: 2

    How can sole number crunching power find the actual oil/gas reservoirs?

    1. Re:So how does it work? by Garin · · Score: 5, Informative

      Seismic imaging. Imagine solving a wave equation (acoustic, elastic, or worse) over a 3D grid many kilometers on a side with grid spacing on the order of meters. Imagine you're doing it with a strong high-order finite-difference code. Calculate for tens of thousands of timesteps. Now repeat that entire thing thousands of times for a given full survey.

      No matter how much computer you have, it's never nearly enough for seismic imaging.

      --
      In any field, find the strangest thing and then explore it. -John Archibald Wheeler
    2. Re:So how does it work? by toastar · · Score: 5, Informative

      Processing Seismic Data takes a ton of power, There are techniques that are well known that we still can't use due to the lack of computer power. The last big advance was RTM(Reverse Time Migration). This was first done on 2D Data in the 80's, But didn't become reasonable to do on 3D's until about 2008-'09. This improvement in imaging is one of the drivers is subsalt exploration. The next big step is FWI (Full Waveform Inversion) We still don't have enough power to run this mainstream yet, The main idea is the stuff we mute out as noise is actually just data that we can migrate back to the original location. The other Item more power helps us with Is running Migrations at higher frequencies. right now we record at 250Hz(125 nyquest) but only process at 60Hz, This is mainly due to the price of computer time. doubling to 120Hz requires 4 times more computer time. But allows us to double our image resolution from to 50 meters to 25 meters. Considering some of our target reservoirs are as narrow as 20 feet, This type of thing is important.

    3. Re:So how does it work? by rwise2112 · · Score: 2

      Seismic imaging. Imagine solving a wave equation (acoustic, elastic, or worse) over a 3D grid many kilometers on a side with grid spacing on the order of meters. Imagine you're doing it with a strong high-order finite-difference code. Calculate for tens of thousands of timesteps. Now repeat that entire thing thousands of times for a given full survey.

      No matter how much computer you have, it's never nearly enough for seismic imaging.

      Exactly right! Geophysics (and particular oil & gas) was one of the primary driving forces around supercomputer design, and computing in general. As an aside: Many of you are probably not aware, but TI started out a geophysical instrument manufacturer.

      --

      "For every expert, there is an equal and opposite expert"
    4. Re:So how does it work? by sdguero · · Score: 2

      Much better explanation than I had! :)

      I used to work for an HPC vendor, about 1/3 of our sales were to oil/gas. From what I understood it was pretty easy for our customers to parrallelize their workload to run across thousand of off the shelf servers with gigabit ethernet vs having to buy a much more expensive setup with high speed interconnects.

    5. Re:So how does it work? by Rising+Ape · · Score: 2

      The other Item more power helps us with Is running Migrations at higher frequencies. right now we record at 250Hz(125 nyquest) but only process at 60Hz, This is mainly due to the price of computer time. doubling to 120Hz requires 4 times more computer time.

      Worse than that, doubling the frequency requires 16 times the computer time, as you need to halve your spatial sampling interval to prevent aliasing. Assuming we're talking about wave equation based techniques such as RTM.

      This is one of those areas which is still very much "no matter how powerful the computer, we'll use it all".

    6. Re:So how does it work? by Shinobi · · Score: 2

      Then you understood it wrong. Seismic imaging is one of those tasks that benefit from SSI etc, and in the absence of that, high-speed&low-latency interconnects such as Infiniband.

      The higher resolution of the grid, the more data you need to pass back and forth, at high speed, with as little latency as possible. Each cell needs to exchange data with neighbours, and the datasets are very large.

      Back when I started out with HPC, this was one of the tasks where a 128 CPU Origin 3000 SSI with 128GiB shared RAM, beat a 400 Xeon Linux cluster using SCI interconnect, despite the Xeon on paper being faster. With gig-E it would have been even worse for the cluster.

  12. there are more of these beasts hiding in the wild by mrbongo · · Score: 5, Interesting

    There are actually quite a few of these big machines. Most of them in Houston, but some in Europe. Every major Oil Company and Every large Seismic company has one. They are all huge, and I have never seen on of them shut down to run benchmarks, and most folks don't externally advertise their existence. The cost too much and they have too much backlog and will never appear on a bullshit benchmark web page reserved for underutilized supercomputers. To the person asking if these are overkill? No, The folks referencing the RTM, FWI etc have hit the equations on the head. One processing job may take 6 + months to run a single migration using 20,000 + cpus. They run all kinds of cpus' gpu's and change out masses of them every time there is a step change in a chip for efficiency. If they had chips 100 times more powerful, they have equations waiting for them. with regards to the person or people talking about carbon ending it all etc.. These machines enable the reservoir engineers to target more reservoirs and then deplete these reservoirs more efficiently leaving less hydrocarbon behind (theoretically reducing the number of dry wells) We will never run out of oil, we will however run out of the technology to efficiently extract it from the ground. ( or it will become cost prohibitive) Carbon use however is another kettle of fish. Making hydrocarbon more expensive will only push coal back front. (look at china, germany etc) Until use is addressed, alternative will be what they could be. Doing things like shooting ourselves in the foot with ethanol is a good way not to proceed though

  13. Re:Most powerful commercial supercomputer by DanielRavenNest · · Score: 2

    It is not like you can take a bitcoin rig and do anything worthwhile with it.

    I'm typing this slashdot comment right now with it. My graphics card is mining in the background, in parallel with providing the display for both monitors. The displays only slow it down about 2%, and playing video on the larger monitor eats about 1/6 of the mining speed. If I was playing a heavy duty video game I would have to shut down the mining program, but most other PC tasks can work just fine in parallel.