I do always discover problems downloading the queue of wating cuda_fermi files.
If i do not restart this queue uncountet times the seti queue on my PC stopps because there no swu's to do left.

This really sucks !

It's not a bandwide-problem, as i am connected through a 50 Mbyte DSL line.

Any idea how i get this solved ?

Any idea how i get this solved ?

Tell Berkeley to give Seti a bigger connection out of their lab, as they are limited to using the 100Mbit connection that's saturated.

As you can see when you click that link, those greens are pegged to the max 93Mbit. Then it doesn't matter what super connection you have, if the project can't get the work out, your connection can't get it in.

---

We are talking about files of 367k each. If they lack the bandwith to transmit such files in one peace at a time i do deeply feel sorry.
Somehow it reminds me to my first 300 baud accustic coppler ages ago.

If calculation on my machines stops for 3/4 a day because of lack of files i don't need to tell them to raise their connection. Iam sure they are aware of this bottleneckß.

Slowly i start to understand why they are bagging for donations as well.
You see, it took me much afford and some bucks as well to get thermic problems on two machines solved, so that they can run 8 cores and one GPU at 100% permanently. Beside all this there is the enhanced power consumption for wich i have to pay as well.

Anyhow i do daubth that some intellgent specis will tranmitt radio signals into nowhwere. Considering the time it will travel it would make such afford senseless.

So if SETI lacks the neccecary bandwiht i may should spend my cputime to some more useful projects ?

So if SETI lacks the neccecary bandwiht i may should spend my cputime to some more useful projects ?

That's how BOINC is set up, if one project can't do it for you, add others. Enough to choose from: http://boinc.berkeley.edu/projects.php.

Oh and yes, the Multibeam tasks may only be 360KB, but when several thousands of computers are at the same time trying to get those tasks, they saturate the bandwidth anyway. Nothing said about saturating the database. When Astropulse tasks are being split, it's even worse, as these are 8MB a task. Have two of them being uploaded to you alone and you saturate the bandwidth already (2 * 8Mbyte * 8bit = 128Mbit, more than there is bandwidth here); extrapolate that to having more than one host download APs and you see it's gridlocked.

---