Folding BigAdv to end January 2015

Good decision for the future of the project, in my opinion.

When I started Folding the ethos was that any contribution was valued, whether it was one core or 100. The points inflation started with SMP and GPU, but at least these were achievable on a moderately decent desktop and the points weren't totally disproportionate.

The BigAdv points are disproportionate. It sends the message that the project may be called Folding@Home, but Stanford's main interest is in attracting 48-core monsters. A computer which can run BigAdv is not a 'home computer' according to any popular conception of the word, and Folding@Home should be about home computers. Essentially it's changed from a community movement towards renting workstations for free, and I don't see that as a positive development.
 
Big news, a 2/3 year phase out would have been nicer to those who have invested in equipment especially. I can see a lot of the heavy users being pretty ****ed and moving to other projects.

I wish they would just make it work with Boinc and make everyones lives easier!
 
Mattus said:
Good decision for the future of the project, in my opinion.

When I started Folding the ethos was that any contribution was valued, whether it was one core or 100. The points inflation started with SMP and GPU, but at least these were achievable on a moderately decent desktop and the points weren't totally disproportionate.

The BigAdv points are disproportionate. It sends the message that the project may be called Folding@Home, but Stanford's main interest is in attracting 48-core monsters. A computer which can run BigAdv is not a 'home computer' according to any popular conception of the word, and Folding@Home should be about home computers. Essentially it's changed from a community movement towards renting workstations for free, and I don't see that as a positive development.
Click to expand...

I totally agree. When people start using up the HP Cloud Computing trial just for folding and can get a million points a day for BA, where's the incentive for the home folders to take part? If you check the number of identified individuals (i.e. not folding as anonymous) and who have returned any units in the last seven days, the project is down to around 23000 individuals. I am certain that that is at least partially down to the disincentive of seeing one computer achieve in a day what might have taken someone with a single home PC achieve in a year.

Folding SMP has become steadily less and less worthwhile to the point where I really only fold on my GPUs.

My issue with it all is that I really can't see a good DC project that uses GPU effectively other than FAH.
 
It seems like a recognition of the limits of distributed computing. The more you focus away from the bulk of home users and towards a small number of people with very powerful machines, the closer you get to a 'traditional' supercomputing model.

The present mining boom can't be helping either. If I had dozens or hundreds of GPUs doing F@H, I'd definitely be tempted to switch them all to Litecoin mining and start accumulating cash rather than points. Shame, since F@H is a very worthwhile project. But like many DC projects it's been managed very badly.
 
Last edited:
They should do the same with GPU computing as well then by that logic because a quad gpu rig costs way more to buy than a quad CPU rig and uses way more electricity and is more expensive to cool etc etc etc.

Don't be fooled this is nothing more than politics and mismanagement.

Oh and the HPC cloud stuff was 2 years ago and you couldn't get much more than 60K out of it if you were an early adopter, and was squashed pretty fast by HP plus pretty heavily frowned upon by the community.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom