I've been using BOINC for years now on my Android Phone, it's App is amazingly well written, but it's missing one feature: Finish on AWS.
The idea came to me recently when I had a Task that was 70% in completion but only hours away from its overdue date. At that moment some part of me would have paid 5$ to see the task finished on an AWS instance so the work of my trusty little phone wasnt for nothing.
It's wasteful and stupid from a personal point of view... But it triggers the same feeling some mobile games trigger + I wouldnt buy virtual crystals but actual scientific data.
I've played with this, once upon a time it used to be a screensaver - you can run this headless now.
The idea to pair up worthy projects (think scientists folding proteins or doing something computationally intensive) with spare cycles is an old one.
I'm not sure how efficient this approach is. I got the impression when I tried to donate some cycles recently (spare servers) that its not widely used. Anybody has any experiences they would like to share? I'd love to hear.
BOINC itself is the white-label open-source[1] tech that other projects use, including SETI@home. I participate in IBM's World Community Grid[2] project, which uses the BOINC client. (it's running on my home machines, including some old/spare Android devices)
The IBM project is definitely active — their homepage currently says there are 721k volunteers participating, and there's a monthly email newsletter that covers their sub-projects' progress.
For an example of a current project, see OpenZika[3].
I see projects like BOINC and Jason Scott's ArchiveTeam Warrior[4] as a sort of "botnet for good" — if you have some bandwidth + margin in your monthly electricity budget, why not? I'm essentially spending a bit of electricity each month to help contribute to (and be a participant in) these projects. Their marketing must be working, because it "feels good" to be a part of them. :)
I don't know enough about CPU use + electrical costs to truly gauge the impact of my devices, e.g. vs. spending the same amount of $ on a fast, modern server running on clean power in an Icelandic datacenter. But having an old laptop or two + a few old Android devices running 24x7 hasn't noticeably appeared on my monthly PG&E bill.
I am wondering if it's more efficient to spend CPU cycles of PCs and mobile phones at home at the cost of increasing the electricity bill or to donate the same amount directly to the project so that they can use it for optimised (GPU?) servers that have better computational power per dollar.
The BOINC client is (afaik) pre-installed on many IBM machines; it's been on both laptops I've been issued. I suspect that IBM is one of the biggest supporters of BOINC.
Disclaimer: IBMer, simply sharing an observation, not talking on behalf of organization, etc.
> I'm not sure how efficient this approach is. I got the impression when I tried to donate some cycles recently (spare servers) that its not widely used. Anybody has any experiences they would like to share? I'd love to hear.
I remember it has been used to crack firmware signing keys for TI calculators.
If memory serves me right, this was added after someone traced down their stolen laptop because the SETI@home screen saver was running on the stolen machine.
The "Choosing BOINC projects" [1] page gives a good overview of what projects are using BOINC and which clients are available on which platforms or can use GPUs.
Yes it is. That's Einstein@Home [1]: "Einstein@Home is a program that uses your computer's idle time to search for gravitational waves from spinning isolated compact objects (among which are pulsars) using data from the LIGO gravitational wave detector."
I was wondering if its coincidence. Recently there's been an increase in activity on the BOINC crypto currency called gridcoin. You can see here: https://www.poloniex.com/exchange#btc_grc
