SETI@home has announced that they will no longer be distributing new work to clients starting on March 31st as they have enough data and want to focus on completing their back-end analysis of the data.
Back around 1999, I brought up a bunch of high-end Unix equipment, after it was moved from another data center to “mine.” This included a Sun E10000 — a “Unix mainframe” — $15M of EMC disk cabinets, a 384-DTL tape-changer robot and a bank of tape drives, and various development machines, like a 20-CPU E4000, and a E450.
The E10K cost $500,000 at the time, and was “half-racked” with 44 CPU’s. This, of course, was the first thing I setup, because it was extremely interesting. So, while everything else was being configured, it was just sitting there doing nothing.
I had heard about SETI@home, and thought it was cool, but I didn’t want to commit any of my personal machines to running it. But I knew that they had created clients for every platform available, including Solaris, and I put two and two together.
I set the application up to run on the E10K. Forty-four instances of it. I figured it was a good “burn-in” test for the config. For about a month, it ran at 100% utilization. Of course, when the Oracle developers finally started using the machine in anger, I shut it all down. But, in that month, I shot up the leaderboards for the project, and cracked the top 1000 contributors.
Very cool to see their link to other “open” distributed science projects you could donate CPU time to.