PcPerf bot

We have added one word to the End User License Agreement (EULA), but it is an important word.  The EULA can be found at

http://folding.stanford.edu/English/License

We added the word "written" to this sentence:

We felt that this was an important addition in order to avoid any confusion.  There have been a few situations where donors felt that they had permission, but the owners of the computers did not.  Having written permission is the best way to make sure that there is no doubt.  It also gives protection to the donor in that he/she would then have proof of permission, avoiding problems involving oral agreements.

Voir l'article complet

We have another paper that came out recently (paper #60 at http://folding.stanford.edu/English/Papers).  This one describes the PS3 client in FAH:

'In this paper, we detail how we were able to get great speed increases for Folding@home (and actually certain molecular dynamics calculations in general) on the PS3. This is our first paper using the PS3, laying out the "how does it work," with a follow up paper in the works describing the results obtained in FAH from PS3 clients. 

It is also worth noting that this paper is a collaboration between FAH team members (Luttmann, Ensign, Vaidyanathan, Houston [now at AMD], Jayachandran, Friedrichs, and Pande) with developers at Sony (Rimon and Øland and their coworkers). 

Here's the more technical abstract as well:

Voir l'article complet

We have another paper (paper #59 at http://folding.stanford.edu/English/Papers)

which recently was published, representing a major step for FAH. 

This one deals with the ribosome. The ribosome is a fascinating molecular machine, responsible for the synthesis of proteins. For this reason it is of fundamental importance to protein folding (as the last step in the central dogma of biology) as well as to human health (since the ribosome is the target of a very large fraction of antibiotics). One of the questions revolving around ribosome function is why is there a large tunnel inside the ribosome, through which proteins exit after being synthesized. In this paper, we used "bigWU" classic clients (clients which allow larger systems to run) since the ribosome is so huge that it would not run on regular classic clients.

 The primary goal of this paper was to analyze the surface of the ribosome tunnel. Understanding the nature of this surface would be useful for both understanding the fundamental nature of protein synthesis as well as how key antibiotics interact with the ribosome. An interesting related discovery was the identification of a potential "ribosome gate" which can open and close selectively, based on what is interacting with the gate. This suggests novel hypotheses for several aspects of ribosome function as well as interesting new directions for work on studying the ribosome and for new routes for antibiotics.

Voir l'article complet

I am very happy to announce that a first key work (paper #58 at http://folding.stanford.edu/English/Papers) from the Folding@home project on Alzheimer's Disease (AD) was just published

Abeta misfolding and aggregation is believed to be the cause of Alzheimer's Disease. Simulations, like Folding@home, are a natural way to understand this process. However, there are several key challenges for simulating the key step -- oligomerization.

This work represents a new way to simulate Abeta oligomerization, with a key advance of being able to simulate experimentally relevant timescales and concentrations, using a novel method. We use this new method and the power provided by Folding@home donors to simulate oligomerization in all-atom detail. This has lead to specific predictions about the process, which we are now testing experimentally. 

In many ways, this paper is the "tip of the iceberg" for the Folding@home activities in AD, with a lot more interesting results to come, especially in terms of experimental tests of our predictions and interesting new possibilities for new drugs and AD therapeutics.  So, while we're excited that this result is now past peer review, we're even more excited for what's coming down the pipeline, waiting peer review.  We'll keep you posted as more results become public, hopefully with some even bigger announcements in 2009.

UPDATE:It was asked which clients participated. This work started several years ago and took some time to analyze and then publish. So, it ran exclusively on classic clients. For the follow up simulations, we are using a mixture of GPU, SMP, and classic clients. Due to the large number of classic clients, they allow us to calculations not possible on the other platforms. However, the raw speed (but smaller number) of the GPU and SMP clients allow us to get a good rough idea quickly, refining later with classic clients.

Voir l'article complet

It's been a while since I've posted on this topic, so I thought it would be a good time to talk about the long term vision for FAH.  In the first few years, we have done a lot to build up methodology (novel distributed computing algorithms, tests of methods), especially in the last years working to bring GPU's, PS3's, and tightly coupled SMP calculations to a distributed computing platform.  While all of this has been going on, there has been a lot of work which would have a direct impact to disease and human health, and donors should see more results (i.e. papers) coming out in this direction soon.  In particular, results on Alzheimer's Disease, Huntington's Disease, and some other surprises in the works (sorry, some details are best left until peer review is done).

In terms of a long term vision for the FAH software, our short term goals is to shore up the SMP client.  Just as there was a big improvement from the GPU1 -> GPU2 client, we have been working steadily on an SMP2 client -- a version which is much easier to use, requires less donor effort, and scales much better.  This really is a combination of a modified client and (especially) new cores.  This effort has been going on for about a year and it's far enough along that I'm starting to talk about it publicly.

In the previous post, I also mentioned about some of our work behind the scenes in revamping the backened of FAH, i.e. a whole new set of server codes.  We have code for the workserver and collection servers, but in time we expect the full backend to be cleanly rewritten from scratch.

Longer term, with a more solid GPU2 and SMP2 clients done, we can then move on to other areas.  Our goal is to be bleeding edge in the nature of our science, but get back to the simplicity and stability found in the classic client.  We still have a ways to go, but I am excited that projects that were started some time ago are now looking like they will see the light of day.

Voir l'article complet

We've made some changes to the Collection Servers (CS) which seems to be making a big difference.  We've migrated them to a new network (the old net had many issues, as discussed in previous posts) and recently made some configuration tweaks.  The two new CS's (171.67.108.17 and 171.67.108.25) seem to be doing reasonably well right now.

Hopefully, this will help donors get WUs in faster and allow the FAH team some leeway when a server goes down, knowing the CS is there as a backup.  We'll continue to monitor the CS issue, since this is such a critical part of FAH.

Voir l'article complet

We got the Main AS back up and all looks good on that box.  The main systems will take some time to get completely back in order (say a couple of hours), but it's looking good so far.

Voir l'article complet

The main AS is down.  The two consequences of this are that the secondary AS is handling all of the load and the main stats update have been suspended until this is resolved.  The points are being kept locally on the servers, so WU's are being credited, just not showing up on our web site.

Voir l'article complet

Engineers at NVIDIA (notably Scott LeGrand) have come up with a theory for the EUE's seen in core 1.15 (and a few others in the 1.15 to 1.18 range) on certain hardware. They found that this core had code optimizations that drove the GPU so hard that it would draw a lot more electricity (one sign of this was running hotter). In some boxes, this was too much electricity and this lead to numerical instabilities. When the same machine was given a beefier power supply, the problem went away.

We've been told that 8800's require 600W power supplies, but we're finding that even a little bigger (eg at least 650W) is important to leave some room for error. We are working to see if there is some way to detect this issue in software, but for now, if you're getting EUE's on the NV GPU client, this is something to consider.

By the way, this will be very important for us to consider future code optimizations. NV core v1.19 removed some optimizations to solve this problem, but there are many cards which would run fine w/this more optimized code. If we can find a way to detect whether the card can draw enough power, we may be able to choose different code paths to allow for greater optimization for cards which can handle it.

We're still looking into this. For now, if you're seeing issues with your card, please consider trying out a bigger power supply. We will continue to look to see if this is indeed the problem and what we can do to help the situation such that the code runs stably on all machines.

PS Please post replies here:http://foldingforum.org/viewtopic.php?f=52&t=6755

Voir l'article complet

We're happy to announce that the PS3 client for Folding@home has won a prestigious design award:

Here are some more links.  First a news site discussing it

http://kotaku.com/5078119/foldinghome-wins...se-design-award

and the Sony Press Release

http://www.scei.co.jp/corporate/release/081106de.html

Voir l'article complet

We reset 171.64.65.83 and now the PS3 clients are looking better.  We'll keep an eye on this, but it is the weekend in California right now, so we are on a limited staff (and thus response will be a slower than on week days).

Voir l'article complet

This will involve brief (3-5 minute) network interruptions during the time period.  Most of FAH will not be involved, but the 171.64.65.XX network will see some downtime.

Voir l'article complet

The Stanford CS building was off line briefly today starting at about 1:30pm pacific time, lasting for about 3 hours.  This meant that all of our machines on the 171.64.65.xx subnet were off line.  It looks like everything is back on line and Stanford IT is looking into causes.  This will likely lead to a bit of an excess load for a while until things settle down.

Voir l'article complet

We will have a brief (1 hour) server outage tomorrow, October 28 at 10am pacific time.  It is expected to last about an hour.  This will only affect machines on the 171.64.65.XX network, so most of FAH will be up.  However, we expect there to be a server backlog for a little bit after the machines come back up.  Also, it's likely that one server or so will have trouble coming back up and will be delayed.

Voir l'article complet

We're continue to work to keep the PS3 servers stoked with jobs.  For those who are having problems and are not running the latest FAH software (based on Life With Playstation or "LWP"), please upgrade.  We need to concentrate our servers on the latest clients and cannot guarantee work for older clients.

Voir l'article complet

We're looking into issues with our PS3 client.  We have a short term fix in place, but are working to do more right now and hopefully will have more to say later today.

Voir l'article complet

We found an issue with the stats server this weekend and have resolved it.  The stats will now be updating again, although it may take a few hours from now to resolve the backlog.

Voir l'article complet

As I've discussed before, we have been having issues with a particular subnet (171.64.122.XX) for Folding@home servers.  The symptoms were dropped connections and general challenges returning WUs.  We asked for a new network, just for Folding@home, and after much red tape we have received it and it has been installed on a few trial servers.  So far, it is looking very promising.  VSP07v (171.67.108.11) is an example of a server interface on the new network and it is behaving very well.  We are in the process of updating additional servers.  It's still possible that this isn't a network issue, but given our results with VSP07v, that looks unlikely, so we hope we can now see the light at the end of the tunnel.

We have also started rolling out new GPU servers.  The GPU serverstat pagehttp://fah-web.stanford.edu/localinfo/contact.GPU.html

now shows more servers in general, as well as a much lighter load.  This was possible now that we have the new network.  If all looks good, we'll plan on rolling more out in time.

Voir l'article complet

We restarted the main AS.  The stats will also skip one cycle due to this, but everything else should be back to normal.

Voir l'article complet

There have been two issues which have been causing FAH some series issues. I wanted to give an update.

First, networking on subnet 171.64.122.XX seems to be very slow and easily overloaded. Our diagnostics (and donor input) points to this issue being caused by a firewall on this subnet that can't handle the load. I have asked 3 different branches of our IT dept looking into this, but nobody has any fixes just yet. Getting impatient with this slow response, some time ago I requested for us to get a new network for our machines. This request is being processed. Once it's accepted, it will take a little time for them to get the new net in there (they may be able to do a VLAN, but more likely they will have to run a new physical cable, since the VLAN would still be behind the firewall). This subnet also has some of our collection servers, so this will be a big help there too once this is resolved.

Second, the current server code can get overloaded. When it does, it slows down. However, new code in the server notices this and restarts the server binary. This leads to downtime of about an hour when this happens. While, this does autofix the problem with just a little downtime, I'd like zero downtime (as would most people). I have paid a professional software house to rewrite our FAH server backend from scratch. That is almost done (it's in QA right now, with some somewhat minor issues to address). This new server code should address this issue (and other issues) with the server code, but may introduce new issues that need to be smoothed over. However, the rewrite is MUCH cleaner architected and so that will be important going forward in the future.

I just wanted to give a bump to let people know where we are. These issues aren't quick to resolve, but we are making progress. The new server code in particular will be a big help in the next 5 years of FAH, due to its rearchitecting and much cleaner code.

Voir l'article complet

We've been looking into the issues some people have been finding with the viewer. On the ATI/AMD side, the catalyst 8.9 drivers seem to have resolved a number of issues with the viewer.

On the NVIDIA side, we have also now released on our download page a special version of the viewer designed for NVIDIA GPU's; this viewer has modifications that makes the viewer run much more smoothly and in general and behave better on a broader range of NVIDIA GPU's. If you're curious to run the viewer, please check it out (on our high performance client download page).

Voir l'article complet

In partnership with Sony, we've rolled out a new update to the Folding@home client for the PS3, now built into their "Life With Playstation" (LWP) application. They've done a great job of building a beautiful piece of software. Folding@home is running underneath in the background. Noam Rimon, the lead developer on the Sony side put it well in an earlier blog post as well as a more recent post on the PS3 blog

We've also updated the science code to do the same functionality of what the GPU2 code can do today. However, people will notice that the ns/day will go significantly down. We want to stress that the points per day remains the same. This means that one gets more points/ns since the calculation is more complex, so don't worry too much if your ns/day is lower. LWP takes up a little extra compute time from the science, but not much. Most of the decrease in ns/day is due to the fact that it's doing a much more complex calculation (which is good for the science side of this).

There's also a nice Video demo done by Noam Rimon showing the functionality of LWP:

We'll give further updates about this as time goes on, but we're excited to have the new science functionality in and to have a client which hopefully can even further broaden the participation of PS3 owners in Folding@home.

PS Some people have noticed that LWP seems slower. In particular, the ns/day count is down, but that's not what it might seem. We're doing more complex science which runs slower on the PS3, so that's responsible for most of the slow down you're seeing.

For those who are interested, it's slower because we're running a more complex version of the GB implicit solvent, which is much more computationally demanding; the upside is that this would be much more accurate. This is also the main solvent model we're running on GPU2 right now. In fact, earlier PS3 results pointed us to the significance and importance of moving to this new model, so this (and GPU2) reflects what we've learned so far.

However, there is a small performance hit due to all the other life with Playstation (LWP) eye candy. However, it's not very big in the end (the Cell and RSX GPU are very powerful), with most of the slow down in terms of ns/day coming from the new science code. The upshot of the LWP platform will hopefully be getting a lot more PS3 donors (but I guess time will tell there).

So far, the response will be strong and I expect that this could in time double the FLOP count from PS3 donors in Folding@home, which itself will be pretty impressive. More importantly, with the advanced science now rolled out on both the PS3 and GPU2 platforms, we're well-poised to tackle some much more complex and interesting calculations.

Voir l'article complet

The PS3 server 171.64.122.73 is down and we are working on it.  However, there are other PS3 servers that are available, so this shouldn't be a major issue.

Voir l'article complet

A server hung last night. It's back running now and the stats are updating again (as of 6:30am PDT). The backlogged points will appear later today.

Voir l'article complet

Our server is back up and we're in the progress of doing a stats update. This update alone will take some time and so WU's will take longer to get through the system for about 24 hours. The good news is that all the key systems are back on line.

Voir l'article complet