4 ms·

BOINC: Compute for Science

22 points by smpetrey 11y ago

threeio 11y ago
Wow... I haven't seen BOINC in a decade.. As I remember SETI@Home converted its project from its custom program over to BOINC back in the day. -- https://en.wikipedia.org/wiki/SETI@home https://en.wikipedia.org/wiki/SETI@home The initial software platform, now referred to as "SETI@home Classic", ran from May 17, 1999 to December 15, 2005. -- So about a decade ;)
ChrisGranger 11y ago
As a longtime BOINC user, I'm sometimes frustrated by how tiny a percentage of our global (idle) processing power is being used by distributed computing projects, especially projects related to medicine. It seems like such a good idea, and I wish we were doing more...
- tedunangst 11y ago
  If every idle process started running simulations full bore, that'd be a lot of heat... I have maybe 100W of computers "online" at the moment. But the heavy load draw would be well over 1000W. Idle time isn't free. The exceptions would be some previous gen consoles like the PS3, which didn't idle efficiently, but only if you assume they're being left on continuously. I used to run folding@home during the winter to help heat my apartment.
- dragonwriter 11y ago
  > As a longtime BOINC user, I'm sometimes frustrated by how tiny a percentage of our global (idle) processing power is being used by distributed computing projects Hasn't there been significant effort over the last decade or so to improve processor efficiency (for both heat output and power use reasons) by reducing power use when idle, meaning that "idle capacity" is much less of a "free" resource than it used to be?
- isla-de-encanta 11y ago
  BOINC should work more on the personal reward in order to get more participants. for example - include some crypto mining into their client - be included in a free version of windows (microsoft will distribute lite version of windows in exchange for contributing your pc power) - be a default "screensaver" in ubuntu
- aries1980 11y ago
  Longtime BOINC contributor here too. The only thing I annoyed by the projects is the WUs and the source code of the individual projects are not open-source. More researchers could be involved if the data would be publicly accessible.
x43b 11y ago
I always wondered, but not enough to do calculations, that if instead of increasing one's electricity bill by $10, one could donate $15 (tax deductible) to an organization to run the computation themselves. Which is more efficient?
- robzyb 11y ago
  Back of the envelope thoughts: Donating money is unlikely to be more efficient. They have to be electron-juice too, and while it might be cheaper for them to buy it, they also have to pay for hardware on top of it. Option A) You spend $10 on electricity and give them X MIPS. Option B) You give them $15, they spend $5 on electricity, and then they have $10 to spend on hardware. I bet that they'd get far less than X MIPS!
  - zornthewise 11y ago
    They might be able to save up money and get more efficient hardware that eventually beats the science/electricity cost of distributed computing.
Forbo 11y ago
In case helping to advance science isn't enough of a reward for you, there's also Gridcoin[0], the cryptocurrency that you mine by contributing processing power to BOINC. [0] https://gridcoin.us/ https://gridcoin.us/
ebcase 11y ago
Good to see BOINC floating by on HN today. I've recently been dusting off some old Mac minis that run Leopard or Snow Leopard (but no newer OS versions), and was pleasantly surprised that older BOINC versions still function! However, not all BOINC projects support all platform architectures. This site[0] lets you see which projects support a specific architecture; only five support PowerPC Macs. [0] http://wuprop.boinc-af.org/results/delai.py?plateforme=mac_ppc&tri=projet&sort=asc&min=&max= http://wuprop.boinc-af.org/results/delai.py?plateforme=mac_p...

22 points by smpetrey 11y ago

Wow... I haven't seen BOINC in a decade.. As I remember SETI@Home converted its project from its custom program over to BOINC back in the day. -- https://en.wikipedia.org/wiki/SETI@home https://en.wikipedia.org/wiki/SETI@home The initial software platform, now referred to as "SETI@home Classic", ran from May 17, 1999 to December 15, 2005. -- So about a decade ;)

ChrisGranger 11y ago

As a longtime BOINC user, I'm sometimes frustrated by how tiny a percentage of our global (idle) processing power is being used by distributed computing projects, especially projects related to medicine. It seems like such a good idea, and I wish we were doing more...

tedunangst 11y ago
If every idle process started running simulations full bore, that'd be a lot of heat... I have maybe 100W of computers "online" at the moment. But the heavy load draw would be well over 1000W. Idle time isn't free. The exceptions would be some previous gen consoles like the PS3, which didn't idle efficiently, but only if you assume they're being left on continuously. I used to run folding@home during the winter to help heat my apartment.
dragonwriter 11y ago
> As a longtime BOINC user, I'm sometimes frustrated by how tiny a percentage of our global (idle) processing power is being used by distributed computing projects Hasn't there been significant effort over the last decade or so to improve processor efficiency (for both heat output and power use reasons) by reducing power use when idle, meaning that "idle capacity" is much less of a "free" resource than it used to be?
isla-de-encanta 11y ago
BOINC should work more on the personal reward in order to get more participants. for example - include some crypto mining into their client - be included in a free version of windows (microsoft will distribute lite version of windows in exchange for contributing your pc power) - be a default "screensaver" in ubuntu
aries1980 11y ago
Longtime BOINC contributor here too. The only thing I annoyed by the projects is the WUs and the source code of the individual projects are not open-source. More researchers could be involved if the data would be publicly accessible.

x43b 11y ago

I always wondered, but not enough to do calculations, that if instead of increasing one's electricity bill by $10, one could donate $15 (tax deductible) to an organization to run the computation themselves. Which is more efficient?

robzyb 11y ago
Back of the envelope thoughts: Donating money is unlikely to be more efficient. They have to be electron-juice too, and while it might be cheaper for them to buy it, they also have to pay for hardware on top of it. Option A) You spend $10 on electricity and give them X MIPS. Option B) You give them $15, they spend $5 on electricity, and then they have $10 to spend on hardware. I bet that they'd get far less than X MIPS!
- zornthewise 11y ago
  They might be able to save up money and get more efficient hardware that eventually beats the science/electricity cost of distributed computing.

Forbo 11y ago

In case helping to advance science isn't enough of a reward for you, there's also Gridcoin[0], the cryptocurrency that you mine by contributing processing power to BOINC. [0] https://gridcoin.us/ https://gridcoin.us/

ebcase 11y ago

Good to see BOINC floating by on HN today. I've recently been dusting off some old Mac minis that run Leopard or Snow Leopard (but no newer OS versions), and was pleasantly surprised that older BOINC versions still function! However, not all BOINC projects support all platform architectures. This site[0] lets you see which projects support a specific architecture; only five support PowerPC Macs. [0] http://wuprop.boinc-af.org/results/delai.py?plateforme=mac_ppc&tri=projet&sort=asc&min=&max= http://wuprop.boinc-af.org/results/delai.py?plateforme=mac_p...