Promotion and advocacy for distributed computing projects

[SzAndrzejewski]

Dear Forum Memebers and F@H staff

I'm an enthusiast of citizen science, I see almost unlimited potential in distributed computing projects.

I work at University of Technology and I want to promote a distributed computing network system at my university and in my home town

How I can encourage officials nad management staff of my university and my city to help to create such a project?

I see some advantages:

- local distributed computing network can be used also for university/city needs (for example predicting and creating noise/air pollution maps) and when there are free powers they can use them for projects like Folding@home
- This could be aa opportunity to integrate academic or local community

But there are also problems and among them, the biggest problem, in my opinion, is a lack of trust in participating in distributed computing projects.

I've encountered fear that networks like F@H can be used to break into your computer and steal data and information from it. How I can answer these accusations?

Can You help me find more advantages and identify other potential fears or problems? With your help, I would be prepared more for argumentation.

[foldy]

One problem is FAH could disturbe users if PCs are also used for other things. Second problem is power usage which needs to get paid. 3rd problem could be that people might think their PCs get damaged over time when running 24/7 which is not true in general but if a PC fails then they say: oh that was FAH's fault.

[SzAndrzejewski]

Yes, there are also issues that need answers.

One problem, this can be solved by a level of power involved in a project. And F@H client has this solution You can point a level of work of your CPU unit

The second problem, are there any forecasts for how much electrical power is taken for calculations?

The third problem, ok how could I counterargument with such an accusation? Can anybody say with 100 percent certainty that a computer failure results from a distributed computing project?

But I think that there could be also positive arguments

[ajm]

The time is right: there is now more processing power in "common" hardware distributed among the general population than can be gathered in data centers. And the Internet is starting to be fast enough. The potential is huge.

If the project is official, it might be possible to link the participation of donors to some fiscal advantages, or cheaper power. That would be a strong incentive.

The security concern is a matter of a certain technical complexity, which can be credibly addressed only by experts, but such a software can be secured. There is not more risk with it than with the general use of the Internet. Who would renounce the Internet?

For the donors, it is a way to "vote with their own resources" by choosing which kind of projects they support. It is a way to strengthen the bases of modern democracy and to broaden the possibilities of citizens to take part in or to orient the decision processes.

[uyaem]

I've encountered fear that networks like F@H can be used to break into your computer and steal data and information from it. How I can answer these accusations?

The same is true for just about any program you install, just people do not realize that.
Downloading arbitrary installers from the internet has always been very common, but people never really think about the consequences. Fortunately, the majority of the internet is "good".

If someone wants to make sure that F@h doesn't do anything that it doesn't advertise:

• Monitor what IPs and ports it connects to and compare it with the 'whitelist' of servers on https://apps.foldingathome.org/serverstats
• Monitor and 'mirror' the data that is sent by any of the FAHclient
• Monitor what files it accesses

And the ultimate argument, for me anyway:
Don't worry about anyone stealing data from you as long as you're volunteering all your data via your smartphone.

[foldy]

If you run an already trusted app like BOINC or FAH then there is low risk.

If PCs are running FAH then they generate heat. That may be an advantage in winter but a disadvantage in summer. Power usage would increase even more when air conditioning needs to remove the heat. A PC running FAH will use ~150 watts on CPU only and additional 50 to 250 watts on GPU depending on type.

[PantherX]

Welcome to the F@H Forum SzAndrzejewski,

Regarding the security of the application, I am aware that the software was verified by AVAST.

If the people that you are advertising to trust these companies then you're off to a great start; Intel, Google, Sony, AMD, and Nvidia... they all have collaborated at some point and in some cases, are still ongoing: https://foldingathome.org/about/partners/
You also have VMWare advertising Folding@Home as an appliance: https://octo.vmware.com/vmware-appliance-fah/
You also have Microsoft advertising Folding@Home in a sandbox here: https://techcommunity.microsoft.com/t5/ ... -p/1285749
You have Nvidia advertising Folding@Home for their GPUs here: https://blogs.nvidia.com/blog/2020/04/0 ... ronavirus/
Scroll through the last two months on Twitter to see how many companies are supporting F@H: https://twitter.com/foldingathome

When it comes to science, there isn't a single university or lab, rather, it's a collaboration of many labs: https://foldingathome.org/about/the-fol ... onsortium/

[Joe_H]

Please also see this statement - https://foldingathome.org/faqs/miscella ... ty-issues/

[SzAndrzejewski]

Thank you for all your answers. I hope they all will help me.

Do you know any example of very large BOINC network in university among students and all the academic staff?

Or a city which created such a network among citizens?

I think that creation of such a network would be o good idea for an innovation grant application that is going to start in my home city

I'm not a computer scientist but I can submit proposals for this grant. I just want to ask a question.

Is there any topic on this forum which explains what kind of hardware needs to be maintained to work as a central computer unit for distributed computing network and what kind of software needs to be implemented in such a network to work stable?

Or maybe you could tell me something more about this? I only got a little knowledge about this issue e. g. from Folding at Home social media (OpenMM)

[PantherX]

...Do you know any example of very large BOINC network in university among students and all the academic staff?...

None that I know off.. might be better to ask at their Forum: https://boinc.berkeley.edu/forum_index.php

...a city which created such a network among citizens?...

We have millions of donors across the globe. I am sure that there are several donors in a single city. The closest that I can find is on a country level: viewtopic.php?f=14&t=35051

...Is there any topic on this forum which explains what kind of hardware needs to be maintained to work as a central computer unit for distributed computing network and what kind of software needs to be implemented in such a network to work stable?...

If you're talking about hosting a WS (Work Server) then you need stable Debian host which has; 32+ CPU, 64GB+ RAM, and 100 TB Storage with SSD caching. Host must have at least 1 Gb/s dedicated internet connection (https://www.youtube.com/watch?v=HaMjPs66cTs). The F@H Team will remotely connect to it to install the software and start monitoring it. Currently, there is sufficient hardware, just constraints on resources and human time.

...maybe you could tell me something more about this? I only got a little knowledge about this issue e. g. from Folding at Home social media (OpenMM)

Not sure what you mean. If you want to learn about what software F@H uses, then it is GROMACS (https://en.wikipedia.org/wiki/GROMACS) and OpenMM (https://en.wikipedia.org/wiki/Molecular ... ng_on_GPUs).