FAH new Ocores with AVX support

Nicolas_orleans

[H]ard DCOTM May 2016
Joined
Oct 7, 2012
Messages
352
Guys,

I know that the poor project management of FAH has driven most of us out of this DC project. However, there is some technical excitment coming in terms of bio computing. In case you haven't checked the thread in the beta section of FF, below the description that was made by Jesse_V.

Ocores are a really new type of core for F@h. It's not similar to FahCore_17 or other cores that you're familiar with. This is a completely new thing, with a new core, a new backend, and eventually a new front-end. You may have to escape the traditional simulation mentality to fully understand what's going on here. The reason that ocores are different is that they are streaming. Traditionally clients download a workunit, process it, and upload the result and then repeat, and each workunit has its own project-run-clone-generation number. Neither of these things are the case here. Now with this core, clients contribute to a simulation stream, their own protein trajectory. When a PG member wants to study a protein or another molecule, with the ocore architecture they first make what's called a "target". This is a base description of the protein. From the target the researcher can then make many "streams". Each stream relates to the target because it's the same protein; it has the same atoms, the same bonds, the same topology, but the difference is that each stream has a different initial set of atomic positions and a different set of initial velocities. For those that are familiar with the Simulation FAQ, a stream is basically a combination of runs and clones, a target is a project, and the generations are out the window. The new streaming design is not only more efficient, but it's supposed to be much easier to work with in the backend.

Another advantage of the streaming core is that deadlines are easier to handle and it's more efficient. Under traditional cores if I drop out of F@h then it takes weeks for my workunit to timeout and get reassigned. A stream core gains efficiency by synchronizing much more frequently. Now the results are returned to Stanford at every frame. This means that they get the results in a more continuous fashion. Also, every hour or so the Stanford servers are also sent "checkpoints", which are larger and more precise than frames but allow a stream to be resumed from that point. So now, if I go offline someone else can quickly pick up my stream where I left off, and there's less than an hour of downtime. Then I come back to the project I'd get a different stream to push forward. I also want to note that the PG researchers can stop and start a stream at any time.

At the moment, proteneer has created just one target, about 200 streams, and 10,000 steps per frame.

Original post is here

Sounds cool in terms of science and tech, and I am pretty excited to finally see some AVX folding code being tested.

The fun part is : in less than one week BA is supposed to go 24 threads with reduced deadlines, and there was "surprinsingly" no communication or forward guidance, and probably another drama is on the way for the last BA folders. Will probably make more noise than this Ocores developement...

For the remaining beta testers

Requirements:

-64-bit Ubuntu 12.04 LTS or later.
-Either a CPU with AVX support or an OpenCL compatible GPU.
-24/7 internet connection.

Download link: http://proteneer.com/siegetank/
 
Hmmm, I will wait and see. This to me sounds like another string to the part time folders bow. Hopefully us remaining full timers can have AVX support in a traditional core as we haqve our systems running 24/7
 
Hmmm, I will wait and see. This to me sounds like another string to the part time folders bow. Hopefully us remaining full timers can have AVX support in a traditional core as we haqve our systems running 24/7

Maybe a bonus for continual run time. So the longer you continually run it the higher bonus you get.
 
Interesting development. Did they ever hire the person who was going to improve project communications? I've been a happy camper having switched to BOINC and GIMPS.
 
They will have to look under a lot of rocks to find someone that would give a rats ass.
 
I did not see any AVX w/ BA listed in requirements.
 
Hmmm, I will wait and see. This to me sounds like another string to the part time folders bow. Hopefully us remaining full timers can have AVX support in a traditional core as we haqve our systems running 24/7

Yes, it's not sure where this path this lead. For sure, it can have application to part time folding as the Chrome app does.

Also, in my understanding, this kind of design could help to speed up folding trajectories - and thus results - when part time and full time folders fold same projects (that will be the case in February 2015 after BA shutdown) : if a part timer stops his/her computer for whatever reason, no need for the deadline to expire before re-assigning, the trajectory can be completed by whomever is connected to the stream.
 
Interesting development. Did they ever hire the person who was going to improve project communications? I've been a happy camper having switched to BOINC and GIMPS.

Yes, a person was hired.

Results : two guest posts on the blog (synthesis of paper abstracts) and that's it. No communication on FF when servers down / stats down / WU shortage, no threads / polls / whatever to have a direct communication with the donors community.
 
Yes, a person was hired.

Results : two guest posts on the blog (synthesis of paper abstracts) and that's it. No communication on FF when servers down / stats down / WU shortage, no threads / polls / whatever to have a direct communication with the donors community.

The bureaucracy is expandjng to meet the needs of the expanding bureaucracy.
 
I don't have much say when it comes to politics or points. I am merely a programmer and definitely I don't want you guys to feel alienated. I can answer some of your questions regarding OCore, since I wrote it and the backend that supports it.

1) Yes OCore supports AVX
2) This is a totally different core and uses a totally different backend. It is designed for 24x7x356 folding. We actually want to encourage people to hold onto a single "WU" for as long as possible. We can probably make a contest of sorts down the road to see who can fold for the longest.
3) We haven't decided on how to do points yet. There will probably be more than a single "points" metric.
4) This core uses OpenMM, and it doesn't scale as well as GROMACS with respect to the number of cores. However, we will try our best to optimize the core so that it can scale linearly with the number of processors.
5) We're working on unity, so that a single project can be completed by CPU cores, OpenCL cores, and possibly CUDA cores down the road.

Anyways, if you have more questions, you can always find me in #fah on irc.freenode.net
 
I don't have much say when it comes to politics or points. I am merely a programmer and definitely I don't want you guys to feel alienated. I can answer some of your questions regarding OCore, since I wrote it and the backend that supports it.

1) Yes OCore supports AVX
2) This is a totally different core and uses a totally different backend. It is designed for 24x7x356 folding. We actually want to encourage people to hold onto a single "WU" for as long as possible. We can probably make a contest of sorts down the road to see who can fold for the longest.
3) We haven't decided on how to do points yet. There will probably be more than a single "points" metric.
4) This core uses OpenMM, and it doesn't scale as well as GROMACS with respect to the number of cores. However, we will try our best to optimize the core so that it can scale linearly with the number of processors.
5) We're working on unity, so that a single project can be completed by CPU cores, OpenCL cores, and possibly CUDA cores down the road.

Anyways, if you have more questions, you can always find me in #fah on irc.freenode.net

Interesting stuff. Thanks.
 
The bureaucracy is expandjng to meet the needs of the expanding bureaucracy.

That's a common problem, the administration at most universities has to much power, and ends up inventing tons of stupid bureaucratic rules to justify it's massive size.
In the end every new employee ends up as another way of feeding the bureaucracy.

On topic:

Is there any news on this core or the new BA requirements ?

My single BA machine seems to be chugging along posting BA units as it always has.

(I do not read FF, there's a limit to the crap I will put up with)
 
> Is there any news on this core or the new BA requirements?

We're working on something cool right now

> My single BA machine seems to be chugging along posting BA units as it always has.

This new core has very little to do with BA in general, we will work on improving our CPU scaling this year so bigadv machines can take advantage of it. (Currently we suck)

> (I do not read FF, there's a limit to the crap I will put up with)

That's why I'm reaching out to you guys.
 
Proteneer, thanks for the information. We hope you do not mistake the tone of dismay here for FAH administration to include the coders, modelers, and others, such as yourself.
 
> Is there any news on this core or the new BA requirements?

We're working on something cool right now

> My single BA machine seems to be chugging along posting BA units as it always has.

This new core has very little to do with BA in general, we will work on improving our CPU scaling this year so bigadv machines can take advantage of it. (Currently we suck)

> (I do not read FF, there's a limit to the crap I will put up with)

That's why I'm reaching out to you guys.

OK, feel free to post a guide or something similar, when you feel it's ready for the total idiot test ;)

EDIT:
Apparently a guide will be overkill, just dl the proper file and run :)
 
Last edited:
Good to see some development happening. Will be interested to see how the points work out.
 
Now that stats are live, interest may rise :)
I will try to draft an unofficial guide in the DC guides section next week, so everybody has same information in one place as people having the time/interest to browse irc logs and FF beta subforum
 
AFAIK they paused development, and switched to other things that needed fixing or had higher priority, new assignment server code, the android app, fixed core 18 on Maxwell and launching core 21 being a few of them.
 
Good luck to him, pity for the project though as he was one of the few that listened to what donors had to say
 
Good luck to him, pity for the project though as he was one of the few that listened to what donors had to say

Indeed this is true, he seemed a nice chap and motivated.
 
If you do a good job, you move up the promotion ladder.

Once you are promoted to the point you stop being useful...
 
That is true.

allegedly there is a new Ax core in development with AVX support - no idea of an ETA though
 
Back
Top