Well we knew that the
Cell processor -- which makes the PS3 a pretty
cheap supercomputer, along with its myriad other roles -- was well suited to the task of crunching numbers for Stanford's
Folding@Home project, but there's no way we could have expected the unbelievable impact made by 35,000 some gamers in only a few days. In what can only be described as a total hijacking of the leaderboard, PS3s are currently accounting for 734 of the 990 teraflops Folding processes at peak capacity; in other words, Cell processors have more than tripled the project's power even though they only account for around 13% of the total machines grinding away at any given time. Now keep in mind that Sony's boxes have only been pitching in
since midweek, and with tonight's Sunday Night Foldathon -- an event which encourages PS3 owners to simultaneously run the app while they sleep -- we should see even more impressive performance as the slumbering masses donate record numbers of cycles. This would also probably be a good time to direct you towards
instructions for joining Team Engadget, as well as to suggest that even though this is primarily PS3-centric, that shouldn't stop other PC-equipped team members and owners of even bigger supercomputers (we're looking at you,
IBM) from participating.
Read - Folding leaderboard [via
blog you like a hurricane]
Read - Foldathon thread [via
PS3 Fanboy]
posted March 25th 2007 4:45PM
Hey guys, this is incase you want to help but don't have a PS3:
http://ps3folding.googlepages.com
Personally I think that everyone who criticizing this adaption of the PS3 should rethink there opinion. First, the folding@home project may benefit you or a loved one at some point in time. Second, the cost of power is a small price to pay. Especially when you consider the benefits of the project. Third, not everyone bought the console for games, I use my PS3 as a media center, backup web server and workstation for my business. Fourth, this project provides an opportunity to actually participate in a volunteer or public service endeavor with out needing to dedicate your own valuable time and labor. Who in their right mind thinks thats a bad idea?
I have been running the program 24/7 since its release last Thursday morning. And it feels great!!!
Actually, Sony is just catching up on what Bill Gates have donated to his foundation from the XBOX 360 revenues.
As far as Nintendo, they are just making more Mario memoribilia.
@js: As the XBOX division has been in the hole $4 billion+, I find it rather difficult to believe any "donations" have come from here.
@oncology researcher: OK, you did say "drug discovery", i.e., a product on the market. But how much "drug discovery" has come from other things like gene mapping or similar? This is research that should be done, because we know so little, and we're trying to figure it all out. Justifying this sort of thing by trying to tie dollar values to it isn't a very scientific or altruistic attitude.
oGMo-
Agreed- folding@home will overall add to the body of scientific knowledge. However, many people assume that it will produce a miraculous cure for a disease within the next few decades or so... which is very very unlikely. Protein folding is an unsolved problem, and I'm not expert enough in the field to be able to judge folding@home versus the other approaches out there. I do know that if a robust method for protein folding is applied, it will have little to no effect on drug discovery. There are too many variables besides active site binding to make an effective human medicine.
I'm not trying to tie this to a dollar value (don't think I'm in this for the money), but instead trying to tie it to alleviating human suffering. I've been working in oncology research for the majority of my career, and I'm very jaded to the promise of miracle cures. It's counterproductive for my funding, but perhaps excess CPU time can be put towards some other use if people want to make an impact... at least be honest in that in all probability anything that comes out of it will only contribute towards basic science, not medicine.
BTW, I assume that by gene mapping, you refer to the novel target identification craze of the '90s. Very little came out of this as far as I know. Some might be buried in proprietary databases of various companies, but nothing has reached the market as of yet. Most efforts that have been published have been abandoned as non-viable.
I've had a PS3 since the launch date and have Resistance and Call of Duty 3.... They got old and there is jack for games for it right now. Up until this folding opportunity came up, it's sat dormant for at least 3 weeks.... at least I can put my 600 bucks to good use.....
Using this screenshot as a benchmark we see that PS3's provide around 0.0043239 TFLOPS per CPU but the question is how much CPU time is being devoted to F@H when the PS3 client is running? Chances are it is just idle time that the other platforms are donating.
Oh and big plus for the Mac users. PPC and Intel combined we are providing 0.0013588 TFLOPS per CPU while Windows users are only providing 0.0009495 TFLOPS. Unsurprisingly Linux has us beat at 0.0016981 TFLOPS per CPU though.
@ Twist
Ah I did glance at the chart too quickly missed that active vs. total. From this information it would seem the prior generation ATI GPU (X19X0) and the newer PS3 are evenly matched but the complete F@H PC with GPU would probably use 60-80% of the PS3's power depending on configuration. That should soon change with the mid-lower end upcoming ATI R600 and current 8X00 nVidia GPUs in stream computing applications like F@H.
Awesome, thats a good sign, ironically the increased polution is always a bad thing with global warming.
Though in the end its pretty cool, and certainly something to make the PS3 worthwhile till the good games come out. I wonder if the othe major game consoles will get similar applications, might be interesting if they do
Nope...
Even a powerhungry PC only uses 214W/168W actual input/output folding@home. A core 2 duo system probably does around 130/100W input/output if that folding@home with an ATI X19X0 series GPU. PS3 power consumption ranges from 150–200 watts during normal use - closer to power hungry overclocked PC these days folding@home - less efficient Sony hype (hey at least its not as hyped up Steve Jobs BS).
See SPCR PSU fundamentals: http://www.silentpcreview.com/article28-page4.html
ATI's X1950 FTW - fastest and most energy efficient folding@home.
Nope...
Even a powerhungry PC only uses 214W/168W actual input/output folding@home. A core 2 duo system probably does around 130/100W input/output if that folding@home with an ATI X19X0 series GPU. PS3 power consumption ranges from 150–200 watts during normal use - closer to power hungry overclocked PC these days folding@home - less efficient Sony hype (hey at least its not as hyped up Steve Jobs BS).
See SPCR PSU fundamentals: http://www.silentpcreview.com/article28-page4.html
ATI's X1950 Pro FTW - fast and most energy efficient folding@home.
@Kev50027 - In case you are wondering
Peak X1950 Pro power consumption - 65.7 watts: http://www.xbitlabs.com/articles/video/display/x1950pro-gf7900gs_6.html
If you built up a micro ATX box with just a graphics card, a single stick of RAM, notebook HDD and CPU, and x1950 Pro, you'd be way ahead of the PS3 and you could built it ultra silent.
Sorry, that's not right. My desktop is a top of the line PC, with 2 GB RAM, 1.15 TB Core 2 Quadro, and 8800 GTX, and it uses far more power at idle than the PS3, let alone running Folding@home.
Again, you people didn't listen to me, you are all wrong. If the PC were to run the same length as the PS3, the PS3 would have done more folding, using less power to do it than the PC. Comparing how much folding was done makes the PS3 more efficient..
oncology researcher -
I used to work down the hall from these guys at Stanford. It's true that improved protein folding in itself is not going to lead to miracle cures, but it is a major component of improved rational drug design, something that all of us want. Molecular dynamics is not simply about active binding sites - it's about understanding conformations well enough to improve things like permeability, etc. We are simply too early in field's development to rigorously tackle those other problems, hence the focus on active sites.
"I do know that if a robust method for protein folding is applied, it will have little to no effect on drug discovery."
Once again, while things like permeability matter, I simply disagree. I think it's overly harsh to say that a robust method for protein folding will not affect drug discovery. There simply are not any robust algoritms, in the way that I think you mean, in existence, so it is extremely premature to pass judgement on their impact. I think you are attributing a much narrower definition to the field than is warranted.
We simply have not had the computer power for more robust models, so of course it has not had an impact on drug design (present technology allows for ~90% accuracy on chains < 100AA, though I'm a few years out of date now.)
I certainly appreciate your jadedness (I left the field myself for different reasons), but I think you're being overly critical.
Kyoto protocol anyone?
keep wasting our resources
It would be soooo funny If the folding at home app was just an excuse to run up the... Sony Timer!!!
Well thats no big problem since the warranty is only 1 year, probably well have playstation IV by then.
Come buy your PS3 in Singapore, its cheaper here..
Should I buy a PS3 and use this for F@H research. I'm thinking charity and possible write-offs for 2007?
- PS3
- percent of electricity bill
Good/bad idea? Either way it's an idea.
Looks like the PS3's contribution was initially overestimated. I've been logging the data out of interest's sake, check out these three values:
Date,PS3 TFlops, PS3 active, PS3 Total, total Tflops, total active, total total
25 Mar 2007 10:44:59,734,29977,35463,990,228959,1984742
25 Mar 2007 11:45:43,576,29999,35714,832,229064,1985109
25 Mar 2007 13:04:49,515,29965,35977,771,229126,1985497
Active PS3s stays about the same, while their contributed TFlops drops by over 40%
Interesting.
The TFLOP number isn't real time. It's based on a calculation of how many work units come in at intervals. The machines are still counted as active (for two days I think) even though they might not be folding at that moment in time.
Well, I've got my PS3 folding overnight
http://www.hardcoreware.net/reviews/review-356-2.htm
Running the math, PS3 uses about 200 Watts. Let's say you run it overnight for 12 hours a day, 365 days/year. That's 2.4 Kilowatt Hours each day, 876 Kilowatt Hours/ year. Probably about $100 in electricity costs to run Folding@Home.
But wait, there's more! Let's say you run air conditioning in the summer. That extra 200 watts that you generate, it will cost you roughly 400 watts to cool it. On the other hand, in the winter it helps heat the house, but not as cheaply as good 'ole gas or oil.
So let's figure
2.4 KwH/day x 180 days (spring & fall)
1.2 KwH/day x 90 days (winter)
7.2 KwH/day x 90 days (summer)
About 1200 KwH/year at 12 hours day. Probably $150/year in electricity (including heating & cooling). Doesn't seem so free to me.
i think this would be a nice slogan for the ps3
save your world, play in ours.
More like,
"Increase greenhouse gas concentrations in the atmosphere at a level that increases dangerous anthropogenic interference with the climate system, play in ours".
http://en.wikipedia.org/wiki/Kyoto_treaty#Objectives
FYI Kev50027: I'm English. I don't follow American politics.
Also, I make up my own mind on matters of environmental impact, because, unlike most Americans, I actually care about the world I live in (and my children will live in and so on).
*this is Nick btw*
"LIfe is a precious thing. But it is the way of life that people die. Whether it is from Cancer or some other disease, it must happen. There is a reason why these things happen, population control. We are already starting to spin out of control with the population. Sure it is great to save a life. But if there are too many people populating the Earth, then we will all suffer from food shortages and other problems."
Yeah, you can take that into account. Death is indeed just a part of the cycle of life. But, sometimes death can't happen. There are times when death can be prevented. What if it is a child? They have their whole life ahead of them. To let them be taken by some sort of disease is just terrible, because it is a predator you can't even see with the naked eye. And it's not even just about saving a life as opposed to death. It could be saving the person from a life of suffering. Alzheimer's has always been a great burden on not just the people it affects, but those around them as well. Seeing that mental deterioration and them not even recognizing you anymore. I can't even begin to imagine how bad that feels in person. So overall, more people need to join this effort to combat diseases.
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&username=PS3
10th position in what, 8 days(of documented stats) ? thats some good going.
this makes me want to buy a ps3
GPU 1st, PS3 second, but only in specialized calculations as mentioned above by other posters:
http://arstechnica.com/news.ars/post/20070326-why-the-playstation-3-owns-the-pc-in-fh.html
Answers to PS3/GPU speed from the project leader of FAH:
"I think the fundamental misunderstanding is that FAH doesn't do just one type of calculation, but does many, many types, and not all types are sped up on all platforms. What makes the GPU clients and PS3 clients so fast is not universal. They do certain types of calculations (implicit solvent) very quickly (20x to 40x faster), but can't do other types of calculations at all (eg explicit solvent, where the SMP client is kicking ass).
We have been pushing multiple technologies in order to take advantage of the best of what's available, using each technology where it's most useful. If we only had GPU's, FAH would fail, as we couldn't do important explicit solvent simulations (eg important for our drug design efforts and folding efforts). If we only had PC clients, we would lose out on what we could do with GPU's.
FLOPS can look impressive, but there's lots more to the story than just how many FLOPS a client can do. In terms of scientific results, I think our papers, results, and awards give some sense of what we've been able to do and the impact we've had so far, but I'm always most excited about where we're going, not just what we've done so far. The future is looking particularly exciting!"
http://forum.folding-community.org/viewtopic.php?p=171543&highlight=#171543"
Keep the Ps3 bashing somewhere else, it isn't the Ps3 owners fault that the 360 over heats, and breaks down.. xD
Anyways, I worked on 2 WU so far.. and I am leaving the Ps3 on this weekend for the marathon run.. :D
So when is this thing going to gain conciousness and kill us all
Here's a video about the PS3 fighting cancer:
http://www.freshcreation.nl/comments.php?id=988_0_1_0_C
i really wish the 360 would do this too. its notlike they dont have the network in place to distribute itm and at teh same clock speed, with three double threaded procs, And a Much larger install base, i think it could do some reall good.