I think DNETC also used ATI and nVidia both, but it has been down for a few weeks or more. Not sure what is going on there.

DNETC is down because is on the police. Charges for illegal software, etc.

In original from OxyOne (Poland language):

Zawody odbędą się w innym terminie. Nie wiem kiedy, ponieważ serwer DNETC@home jest na policji. Przeciwko mnie postawiono zarzut posiadania nielegalnego oprogramowania itp.
Posiadane przeze mnie komputery to: DNETC@home - debian, mysql, apache, php :)
Notebook ASUS - LMDE - Linux Mint Debian Edition służył do pracy.
Komuś bardzo przeszkadza projekt DNETC. Czas jaki potrzeba na powrót serwera nie jest mi bliżej znany - ale wróci...

I like this project too, but I mean that it dead :(

Thanks for the translation. This is good to know.

You might try what I'm currently trying - learn the relevant computer languages and choose which BOINC projects to help with programming.

Those of you with a military background will remember that the first step in an operations plan is an Estimate of the Situation. I have been following this thread with interest, but the comment that WCG could jump light years ahead if we only had a brain . . . er-r-r, I mean GPU project of course . . . has convinced me that I should post my entirely unofficial and personal Estimate of the Situation.

You already know that HCC has been programming a GPU version. We are in the earliest stages of Alpha testing. [Please remember - there is no guarantee that we will be able to move into Beta testing. We might find a critical bug and have to send it back for more debugging. ] Knreed says that BOINC 6.14 would make supporting GPUs running on OpenCL much easier, but that he can overcome the difficulties. But another comment he made is worrisome [at least to me]. He says that none of his personal GPU cards will run it. To me, this means that only a few members will be weird enough (I mean sagacious enough) to have the correct GPU cards to run the HCC GPU project. (We have a lot to learn about programming to use many types of cards. Our initial project will teach us a lot about this.) So we had better be thinking about progressing light-seconds rather than light years at first.

But these are the sort of difficulties that we have to expect as we get our toes wet in GPU computing. Each year it will get easier to program GPUs and more grad students will be able to write GPU programs. If WCG shows a successful history with GPUs, this will encourage prospective projects to consider supporting GPUs right from the start. If we can gather a critical mass of members with GPUs, we might even get some projects written only for GPUs, saving the effort needed to write 2 different programs to do the same thing.

But this is a long multi-year slog. We are working on it, but I expect no sudden bursts of fireworks. This is more like growing an azalea seedling. Eventually we will have a mature bush producing large flowers.

Lawrence