13 replies to this topic

[Cougar]

Folding@Home lets you donate your unused GPU and CPU resources for disease research. I’ve just started to use it after they added coronavirus projects. You can create teams to keep track of how many cycles you’ve donated collectively; perhaps if others are interested we could create an IMG team.

[Matt Diamond]

I tried it. Unfortunately they don't support Mac GPUs, its CPU only. And there is no guarantee your machine will be used for anything. My 2018 Mac Mini was assigned a (non-coronavirus) job when I got started, but after it completed no new jobs were available in the queue. Apparently different projects write their jobs for different CPUs so there is no guarantee your computer will be in demand.

I had to go into their forums to get this info. When someone with a Mac expressed an interest in the project recently, two members (one a moderator) got snotty about Macs, the usual crap. So I uninstalled the client.

If you have a Linux or Windows machine with an nVidia card, by all means give it a go.

Otherwise, BOINC is a good alternative to folding@home, many projects to choose from, easier setup. I chose the World Community Grid project, and it has been assigning  microbiome immunity and cancer tasks to my machine non-stop. Maybe not as sexy as folding@home, but you do what you can.

[Thain Esh Kelch]

Are there any actual studies that have used these computations for anything? Ie. where can i read about it? Its not just some cover for mining bitcoins now?

[Matt Diamond]

Thain Esh Kelch, on 15 March 2020 - 05:11 PM, said:

Are there any actual studies that have used these computations for anything? Ie. where can i read about it? Its not just some cover for mining bitcoins now?

That thought had occurred!

Folding has a lengthy page of the peer-reviewed published papers that cite distributed work done: https://foldingathom...papers-results/
Now, it is possible to put out lightweight papers on the internet or in sketchy journals, but folding is pretty high profile and I think someone would have called them on it by now.

For BOINC I specifically chose the World Community Grid project because it is run by IBM. Big companies aren't perfect but they do like to protect their brand name so I figure there is likely some due diligence there.

Short answer: who knows. :-) It wouldn't shock me but I think it would be a smaller project trying to fly under the radar.

[Thain Esh Kelch]

Matt Diamond, on 15 March 2020 - 09:05 PM, said:

That thought had occurred!

Folding has a lengthy page of the peer-reviewed published papers that cite distributed work done: https://foldingathom...papers-results/
Now, it is possible to put out lightweight papers on the internet or in sketchy journals, but folding is pretty high profile and I think someone would have called them on it by now.

Thanks. That looks quite fine actually, so I am well satisfied!

[nick68k]

Matt Diamond, on 15 March 2020 - 09:05 PM, said:

That thought had occurred!

Folding has a lengthy page of the peer-reviewed published papers that cite distributed work done: https://foldingathom...papers-results/
Now, it is possible to put out lightweight papers on the internet or in sketchy journals, but folding is pretty high profile and I think someone would have called them on it by now.

For BOINC I specifically chose the World Community Grid project because it is run by IBM. Big companies aren't perfect but they do like to protect their brand name so I figure there is likely some due diligence there.

Short answer: who knows. :-) It wouldn't shock me but I think it would be a smaller project trying to fly under the radar.

I can contribute a little here. So, by day, I'm a postdoctoral biochemist and senior research associate in academia. I haven't taken part in F@H, but scanning the journals being used for dissemination of the results is encouraging. Acc Chem Res and Cell Reports are excellent, serious journals, which are generally very difficult to get results published in.

A simple (although not foolproof) guide is the journal 'impact factor', which is a statistical inference based on the average number of citations a paper acquires, averaged over the journal for a year. Anything over 5 for a journal is good, and anything over 10 is outstanding. Usually your institution will encourage you to submit manuscripts to journals with IF > 5.0 (anything less may not be counted when it comes to promotion etc.). That said, it's not a perfect scheme, and there are many good journals with IF in the 2-3 range.

[Matt Diamond]

nick68k, on 16 March 2020 - 09:08 AM, said:

by day, I'm a postdoctoral biochemist and senior research associate in academia.

.. and by night, you fight crime.

No, no; there's no need to deny it. We can read between the lines! Your secret is safe with us.

[nick68k]

Matt Diamond, on 16 March 2020 - 01:12 PM, said:

.. and by night, you fight crime.

No, no; there's no need to deny it. We can read between the lines! Your secret is safe with us.

In between fighting crime at night I'm also trying to get WoW to run reasonably. Never a dull moment.

[Matt Diamond]

nick68k, on 16 March 2020 - 02:42 PM, said:

I'm also trying to get WoW to run reasonably.

That's 'cause it's a crime how slow it runs! Ha?

Is this thing on?

[Frost]

Matt Diamond, on 16 March 2020 - 06:28 PM, said:

Is this thing on?

God damnit, now Four Little Diamonds is stuck in my head.

[Matt Diamond]

Frost, on 17 March 2020 - 09:51 PM, said:

God damnit, now Four Little Diamonds is stuck in my head.

Now its stuck in mine too.

Keep at least 6 feet away from that link, people. Avoid contact with anyone the earworm has already infected-- it can be transmitted through the air if they start humming or singing it.

[Tibur]

nick68k, on 16 March 2020 - 02:42 PM, said:

In between fighting crime at night I'm also trying to get WoW to run reasonably. Never a dull moment.

Let me know when it does. Those achievements won't earn themselves!

[Matt Diamond]

World Community Grid is adding tasks for pandemic research soon, incl COVID-19. https://www.ibm.org/OpenPandemics
Again, Macs can easily donate CPU cycles to these projects using the BOINC client.

I was going to create an IMG team, but one of our own here has a WCG team already so I joined that.  You don't need his permission to join it but I'll let him decide whether to post his team link.
You can also do it without a team of course.

[macdude22]

https://www.worldcom...mId=J2914ZF75Z1