Slashdot Mirror


China, Russia, U.S. To Build 100MBps Network

prostoalex writes "Gloriad (Global Ring Network for Advanced Applications Development), a scientific data network, will unite academic institutions in China, Russia and the United States with a 100 MBps link. National Center for Supercomputing Applications received a $2.8 mln grant from NSF, and both Russia and China will match this amount to contribute to network build-up. Later this year, as the Associated Press article notes, a new plan will be launched to move the international network to 10 GBps capacity."

13 of 213 comments (clear)

  1. Re:Bit 'B' or little 'b'? by Dreadlord · · Score: 3, Informative

    TFA says its "155 million bytes per second."

    --
    The IT section color scheme sucks.
  2. Mirror, just in case by RickyRay · · Score: 2, Informative

    New Network to Link U.S., Russia, China
    Dec 23, 10:10 AM (ET)
    By JIM PAUL

    CHAMPAIGN, Ill. (AP) - Soon scientists in the United States, China and Russia will be able to collaborate in cyberspace over a new high-speed computer network that includes the first direct computer link across the Russia-China border, developers say.

    The network, expected to go online next month, will ring the Northern Hemisphere, connecting computers in Chicago with machines in Amsterdam, Moscow, Siberia, Beijing and Hong Kong before hooking up with Chicago again, said Greg Cole of the National Center for Supercomputing Applications, one of the leaders of the Little GLORIAD project. Data will flow at 155 million bytes per second.

    "This new network permits us to learn more from each other in areas where we have not worked together in the past," Cole said Monday.

    The NCSA, based at the University of Illinois' Urbana-Champaign campus, received $2.8 million from the National Science Foundation to fund the U.S. portion of the network for the next three years. Russia and China are spending similar amounts, Cole said.

    "As we aim to strengthen our nations' capabilities in research, we also aim to contribute to the cumulative knowledge that lifts the prospects of people everywhere," NSF director Rita Colwell said in a statement announcing the plans.

    The NSF's program officer for the project, William Y. Chang, did not immediately return a phone call to his Arlington, Va., office Monday.

    Scientists have always had computer networks separate from the consumer Internet that assure them the capacity to transfer huge volumes of information at speeds much faster than typical Internet transfers and for real-time collaboration on high-tech experiments, Cole said.

    Little GLORIAD - an acronym for Global Ring Network for Advanced Application Development - will allow scientists and educational researchers to work together on such issues as responding to natural disasters, safeguarding nuclear material, monitoring earthquakes or joint space exploration.

    They also could collaborate to remotely monitor or control high-tech equipment and even could get together face-to-face by video conferencing over the network, he said.

    "This is specifically so our scientists and educators can work together more easily," Cole said. "The technology is really rather amazing with what it allows us to do on a daily basis."

    The fiber optic connection between China and Russia that makes the network possible was completed a few months ago, Cole said. Final touches are being put on the China-Russia link, and the global network should see its first traffic on Jan. 5.

    A formal launching ceremony is planned for Jan. 12 in Beijing, he said.

    Scientists from Russia and the United States have had direct computer linkage for about five years, while Russia and China often exchanged scientific information by meeting in Chicago, Cole said. The new network should strengthen the collaboration between those countries, he said.

    Little GLORIAD is a "first big step" toward development of the higher-speed GLORIAD, Cole said. That effort, expected to be launched later this year, will move data at 10 gigabytes per second, 60 times faster than the Little GLORIAD.

    Computer connections have fostered scientific collaborations that otherwise might not have happened, Cole said.

    "There's some advantage to having people being able to talk more regularly," he said. "There are fewer misunderstandings. I think these networks are going to be more important to the more critical issues that we're all addressing together."

    ---
    On the Net:

    GLORIAD: http://www.gloriad.org

    National Center for Supercomputing Applications: http://www.ncsa.uiuc.edu

    National Science Foundation: http://www.nsf.gov

  3. DWDM & OTDM by Ex-MislTech · · Score: 5, Informative

    With some of the newer Telecom technologies they could hit
    speads of 40 Giga-bits per second if they wanted, most
    likely faster as my knowledge is somewhat dated, ie. 2001 .

    I know Nortel was working on sending 160 Tera-bits down a
    single strand of fiber, and I have seen working gear that
    pushes 40 Giga-bits 2 years ago .

    Here is a article from 1999 that said they hit 1.6 Tera :

    http://www.findarticles.com/cf_dls/m0CGC/19_25/5 46 73084/p1/article.jhtml

    There is now 10 Giga-bit Ethernet ...

    www.10gea.org

    The Telecom links always outpace the current Ethernet high end
    by usually a sizeable amount .

    Peace,
    Ex-MislTech

    --
    google "32 trillion offshore needs IRS attention"
    1. Re:DWDM & OTDM by Anonymous Coward · · Score: 1, Informative

      These are multiplexed data rates, NOT routed. This is done by purely Layer 1 equipment such as digital cross connects and add/drop multiplexers.
      There is no computation involved or route lookup. The source cannot change the destination of the data it is sending so what you have is essentially a lot of point to point connections all close to each other. The individual connections taken together are not a network.

    2. Re:DWDM & OTDM by Ex-MislTech · · Score: 2, Informative

      There are now Optically routed networks .

      http://www.eurescom.de/~public-seminars/1998/OAD M/ Proceedings/Paper16.html

      See "Section 4.0 : Conclusions" at the bottom of the above
      mentioned webpage .

      Routing can be done electrically in the SDH path layers, or optically in the OC layer. Electrical routing is a logical operation, while optical routing is physically implemented.

      Optical routing provides flexibility. It has a coarse granularity (e.g.: STM-16) and avoids using huge DXCs in high-capacity networks.

      In general, DXCs will always be used, since finer granularity routing of traffic (VC-4 and VC-12 level) will be needed.

      Peace,
      Ex-MislTech

      --
      google "32 trillion offshore needs IRS attention"
  4. Re:100 MBPS... by jerde · · Score: 2, Informative

    Be careful with your spelling:

    MBps == MegaBytes per second

    Mbps == Megabits per second.

    100BASE-T ethernet is 100Mbps. Note the small "b".

    - Peter

    --
    INsigNIFICANT
  5. Please, RTA by bugbread · · Score: 2, Informative

    I'm skipping the "F" because I don't want to come off as too much of a heavy, but as long as people don't read the article this discussion runs the risk of being completely off topic.

    This is not a 100 Mbps (or MBps) connection to the internet. This is a private WAN between the connected institutions.

    That means, unless you work or attend one of those institutions, no spam, no mp3s, no pron, no blocking of websites, nothing.

  6. Re:Bit 'B' or little 'b'? by AKnightCowboy · · Score: 5, Informative
    Capital B is not necessarily a standard for "Bytes".. people just like to pretend it is.

    Of course capital B is a standard for Bytes. People are just ignorant to the fact. It doesn't make it any less true.

  7. Re:100 MBPS... by jerde · · Score: 4, Informative

    Oh. Except in this case, the article itself is wrong.

    Stupid, stupid article. Stupider /. editors.

    The network is just a 155Mbps -- that's Megabits per second -- network. That's just an OC3.

    Look at the google cache of a powerpoint discussing this network.

    So this breaks no speed records -- but it is a nice fat pipe into some places that have very limited bandwidth to the outside world.

    - Peter

    --
    INsigNIFICANT
  8. Re:Bit 'B' or little 'b'? by jerde · · Score: 2, Informative

    Except that the article is wrong. Google around for GLORIAD, and you'll find many references to its present 155Mbps speed. (That's a common speed, used by OC3 etc)

    But what's an order of magnitude between friends? :)

    - Peter

    --
    INsigNIFICANT
  9. Re:where's the EU and other Euro nations? by thorgil · · Score: 2, Informative

    Is 10 Gbit/s good enough?

    Check out nordunet:
    http://www.nordu.net/map_nordunet.png

    All universities (and many museums and institutions) in sweden has a dual redundant 2.5 Gbit/s (10 Gbit/s) connection to SUNET which connect to NORDUnet

    --
    Warning: This sig contains a small bug. ==> *
  10. Re:Bit 'B' or little 'b'? by ergo98 · · Score: 2, Informative

    All factual up until this turd...

    Capital B is not necessarily a standard for "Bytes".. people just like to pretend it is.

    Capital B is entirely a standard for bytes. I buy a 256MB DIMM, not a 2048Mb DIMM.

  11. Re:Politically odd? (sorry, OT) by wik · · Score: 3, Informative

    Some more info here...

    http://www.nytimes.com/2003/12/23/international/ as ia/23BEIJ.html

    --
    / \
    \ / ASCII ribbon campaign for peace
    x
    / \