Questions

[Ricorocks]

Are more WU's completed now as compared, to previous influx of new donors? That would appear to be a reasonable assumption, as WU's are in short supply.

Does the F@H team, monitor how many new donors per year & routinely add servers, to meet demand, management? Can/does F@H outsource server demand,

renting while an expense, could help meet demand. Too help manage demand, why not temporarily, remove the client download, from the web site, make

client available to server need/capacity.

[HaloJones]

All good questions.

Adding servers is a matter of resources - money, people - and the money available is not unlimited. The sudden influx where the number of folders increased by a factor of twenty was not an increase that the team had the budget to cater for and it required additional servers being donated to make the upload/download process more smooth.

But the other factor in having so many extra folders was generating the work for the folders to do and that is done by the scientists who can't magically increase the quantity of science so quickly.

Renting servers again is a budget issue which they simply do not have available.

Your suggestion to remove the download is an interesting one; I have no idea if it's ever been considered to effectively put up a "no thanks, we've got enough" message.

[Neil-B]

I believe the concept of removing the download link was mentioned early on in the "surge" - I believe the team will have been aware of that option - I can only guess that they made a judgment call on their ability to respond, which that have done incredibly well - maybe not perfectly but it can actually be easier to expand infrastructure by allowing it to be overloaded and fix/expand to meet the load than to put off the load until you have guessed what needs to expand?

[iceman1992]

Are more WU's completed now as compared, to previous influx of new donors? That would appear to be a reasonable assumption, as WU's are in short supply.

This should answer your question => https://twitter.com/toreler/status/1250022287744675840

[Nathan_P]

another point on server space, given Covid19 is a global problem, any datacentre/cloud provider would suffer a PR nightmare if they refused to help. So far AWS, Azure, Avast, Linus Media group (LTT), one of the forum mod's and Oracle have all provided server capacity - i do not know what financial transactions have occured to make these servers possible, at least some have been donated free of charge

[Endgame124]

Are more WU's completed now as compared, to previous influx of new donors? That would appear to be a reasonable assumption, as WU's are in short supply.

This should answer your question => https://twitter.com/toreler/status/1250022287744675840

Impressive. Are servers still there bottleneck, or is it the science? Throw some non covid jobs in the mix if there isn’t enough covid work to keep the servers busy...

[Neil-B]

If science is part of the hold up, it may well not be the subject matter, but the limited number of researchers who understand enough about folding to actually specify projects - and they can't be magic'd into existence overnight unfortunately

[bruce]

Both.

Outfitting a new server and interconnecting it with the backend infrastructure that FAH was using before the "surge" is not a trivial process but FAH is finding the human resources to do that. We have had offers to help from folks with both large and small organizations (and I, too, do not know the rental costs, if any). Our ability to accept all the offers limits the rate at which we can grow but it's a bit of a supply/demand issue, as well. New donors keep joining at a high rate (see the graph linked in the previous post, and note the flat segment at the far left side of that graph).

On the other hand, science is limited by the time taken to devise, prepare, and ultimately reduce the data from the completed WUs. Scientists from cooperating institutions are also joining forces and contributing their expertise to the project. In fact, I've been helping with the development and testing of some new software that will facilitate research done by the covid19 moonshot team.

[Ricorocks]

Hi Guys,
Not interested in your links, simple yes/no will suffice, more or less WU's returned before the surge, or now with surge. Got it! FYI I was ranked #299 in your ranking, now it appears during pandemic, Stanford is duh wait!
So WHAT SITUATION CONTRIBUTED MOST TO SCIENCE , MEASURED BY WU'S. NOW OR BEFORE?

IF fAH IS WORKING, so hard, how is this money been spent? Do you consult, or even have management in place. To deal with issues like this? Is Stanford/Fah's position golly ghee, shucks, I don't know?

Perhaps F@H should post budget sheet, seeing as how dependent you people are to donors! LOL You folks are traditionally not very transparent, or lag behind informing donors. Perhaps the inverse of too many donors, would open some eyes, & transparency!

Shucks it will cost money, welcome to the world, so far F@H has not manged this well, unless your getting more WU's than normal; if so it's screw the dedicated donors

[BobWilliams757]

Ricorocks,

Reading some of the links, as well as any FOH news lately, would have answered your questions. Per the one link above FOH has done as many WU's in a week as was done in 1/3 of the 2019 work units. That's a lot of WU's getting done. You would also find that a great deal of resources have been donated free of charge. So the infrastructure has been improved quite a bit due to the surge. Can the researchers and scientists keep up? I don't know.

Another recent news article stated that the surge has produced so much help from new donors that FOH has reached somewhere around 2.4 ExaFLOPS which bring it up to being as fast as the worlds top 500 supercomputers. The article went on to state that the COVID-19 research now has more computer power behind it than any other event in human history.

So the short version is, a lot more WU's are getting done. Along with more donors, both in form of financial/physical donations, and the computer resources we are all donating.


As for your long term contributions and high rankings, thanks for that long term support. I'm sure it sucks to have loads of WU's available all the time and then have a surge that takes that all away for a while. But as someone who has folded a lot less time than it seems you have, even back in the infancy of folding when I first did it, the idea was to advance science, not create a playground for the power computer users. And though I'm not as well versed on the overall history of FOH as many, I think they have always just tried to meet the demand of the donors and grow as the community did. I'm kind of shocked that a long term contributor is suddenly so upset and caustic in your respond, considering how large the surge in interest was. Considering a 20 fold increase in people folding, I think they did a pretty good job myself.

[Ricorocks]

As stated in post #1 I assumed, WU's increased do to the surge. Your correct it's all about the science, I've folded when folding was, let's say not as popular. Transparency between donors/F@H has also been a problem for quite sometime. I do understand, caught off guard, by the surge, & the scramble to accommodate.

Does F@h have a plan for the likely mass exit of donors post pandemic?

It seems the surge of new donors, could have been 'managed' better.

Sorry I can't find F@H budget, or management server/donor numbers here

[BobWilliams757]

As for the surge itself, I don't think anyone could really be properly prepared, since much like the virus itself, nobody knew how far and fast FAH would spread. In that respect, I think they have done the best they can.

As for budget/numbers, etc... I have no idea. I was always under the impression that there was no one set group that controlled all the assets, so the budget numbers would be spread out among various colleges, labs, etc.


As for an exit plan, I would hope that they have looked into what is short term vs long term commitments from the new donated hardware, and how it can be scaled based on where things go after the COVID surge ends. I just hope that long term the donations from both the tech company and folder side stay on board at a higher percentage than many are expecting. If even 10% stay on board long term, it would have still probably doubled the WU capacity. And if nothing else, this entire surge will prepare them for the next round, and after the fact analysis of how things went down might teach them some valuable lessons to be prepared for that next round.

[Joe_H]

Does F@h have a plan for the likely mass exit of donors post pandemic?

It seems the surge of new donors, could have been 'managed' better.

Sorry I can't find F@H budget, or management server/donor numbers here

Well, this is not the place to ask such questions. The forum is run by volunteers, most of the answers are from volunteers, and occasionally a member of the F@h Consortium posts and or answers a post. Mostly they post here in connection with Beta testing of new projects, or problems that show up afterwards once they come out of Beta.

Members of the consortium do post on Twitter and the Facebook page. In addition there was a "Fireside chat" with members recently, there is a recording of that if you want to see what they said.

Finally, you keep referring to Stanford. This is old news, but F@h is no longer based out of Stanford, and has not been for 2 years. There are a few people still at Stanford that are connected to F@h, but the Director is now located at WUSTL, with the heads of the Voelz Lab at Temple and the Chodera Lab at MSKCC working with him.

[ajm]

At the beginning of this year, talking of "management" would have been a bit preposterous, I suppose. But now that FAH is (by far) the most powerful supercomputer on Earth, the question can be legitimately raised. I would even consider asking McKinsey if they wouldn't help define a stronger managerial concept, or something. As a courtesy of course.

[HaloJones]

At the beginning of this year, talking of "management" would have been a bit preposterous, I suppose. But now that FAH is (by far) the most powerful supercomputer on Earth, the question can be legitimately raised. I would even consider asking McKinsey if they wouldn't help define a stronger managerial concept, or something. As a courtesy of course.

Hmm. not sure I would ask Mckinsey for help in something that actually has to deliver results but asking them not to help define a stronger managerial concept seems like a good place to start