Slashdot Mirror

← Back to Stories (view on slashdot.org)

Imagine a Beowulf Cluster of Penguin Computers

Posted by michael on from the all-jokes-aside dept.

Pii writes "News.com is running a story about Penguin Computing acquiring Scyld Computing, a company founded by Donald Becker, of linux ethernet driver and Beowulf cluster fame. Becker will stay on as Penguin's Chief Technology Officer, and the companies claim they don't expect any layoffs as a result of the merger."

5 of 179 comments (clear)

  1. I f*king LOVE Penguin Computing!! by smitty45 · · Score: 5, Informative

    those guys are great, and congrats to them. I have fond memories of walking down to their office on Mission St, and carrying one of their new 2U boxes down to where I worked. When we opened the box, not only did we find a machine that is still running right now (over 2 years) but a couple of cool tshirts.

  2. Re:lay-offs by gmack · · Score: 3, Informative

    divisions? The 2 companies combined only amount to 35 people.

  3. Re:It's always the cheapskates who complain. by deadsaijinx* · · Score: 4, Informative

    Hey, Tulip cards and RealTek 81xxs are exactly what I'm running, and I've had no problems with my network, not even under linux. Scyld has done an excellent job implimenting them, and the hardware isn't total crap. It functions perfectly if you know what the hell you are doing with them.

    --
    YOU SUCK BALLS!
  4. No, Scyld will remain in Annapolis by becker · · Score: 2, Informative

    We have moved to different offices in Annapolis about a mile away from our original location. The view isn't as nice, but the new office space has a machine room that is about three times the size of the small office that we previously used.

  5. Re:Anyone actually use a beowolf cluster? by sjames · · Score: 2, Informative

    They're definatly being used for things that are important. They're also well represented in the Top500 list.

    It's worth noting that the super-powerful single machines all use multiple CPUs as well, it's just that their interconnect hardware is different (generally custom, proprietary, and expensive).

    There is nothing you can do on a Beowulf that absolutely can't be done on the custom machines, but the custom machines are generally at least 10 times more expensive. There are problems that the custom machines can tackle that you can't do on a cluster (yet).

    The real question is more along the lines of can you afford to get the custom machine, and will the factor of 10 price difference buy you anything . (Sometimes, the answer to both is yes). In that sense, if budget is limited (and whose isn't?), it may be that they can by virtue of actually being able to afford the cluster at all, but not the single machine.