SCE Joins Stanford's Folding@home Program (B3D ID=32377)

I don't see 1Petaflop to be a problem for PS3s at this rate.

Going on how fast the PS3's teraflop number is going up I don't think it will take that long either.

4hvump2.jpg



WOW!
 
By the way it is being said on the folding forums that the PS3 client DO have the same 'limited type of WUs' restriction that the GPU client does. So from a $/output perspective, plain old ATI GPUs are still the way to go, as I mentioned earlier in this topic.

Judging by the endless discussion of points per WU, tflops, and WU types, Stanford is not providing a decent metric for "value" of a unit of work. They have already stated that points awarded do not accurately reflect "science" done.
Personally I think they are over their heads with the diversity of users, clients and platforms and are just trying to keep participants tuned in and working without actually being able to feed back to the community enough information for it to decide what the cheapest cruncher-per-dollar actually is. Is it a multi-core rig, an ATI gpu, or a PS3? right now this appears to be impossible to calculate .. not because we aren't given vital info, but because nobody really knows..
Weird form for logic.

If I didn't know better I'd say you were grasping blindly for something. :)
 
Weird form for logic.
Weird english!

If I didn't know better I'd say you were grasping blindly for something. :)

good thing you know what I'm getting at, then.

The point I was trying to make isn't that complicated: the metrics for WU points awarded, "science done", and so on, are all over the place. People are making comparisons about platform vs platform, others are trying to work out what the cheapest forum compute platform is per WU, but these calculations are nearly impossible to make.

I'm not grasping for anything, I just think this project, like many DC projects, is pretty shaky. The fewer people who won't even consider any skeptical position on it, who consider it untouchable, beyond criticism because it mentions "cancer" and "cure", the more sure I become.
 
I'm still trying to find where anyone who actually knows the technical details of the PS3 client confirms limited WUs there.
If there was such a confirmation I bet that Onlooker1 is the kind of person who would have slammed it into our faces (just a feeling).
 
The fewer people who won't even consider any skeptical position on it, who consider it untouchable, beyond criticism because it mentions "cancer" and "cure", the more sure I become.
It's not beyiond criticism of course but it has to be objective criticism; criticism made with a degree of understanding of what this project can actually hope to accomplish. Otherwise criticism just turns into babble and mindless monkey poo-flinging.

Since you have avoided answering what qualifications in this field you possess I have to conclude you have none and are thus not actually able to accurately judge the scientific worthiness of folding @ home.

It may be worth noting it doesn't claim to cure cancer. Iyt's not even about curing cancer. It's about understanding the reason and mechanism of why (soem) cancers and other diseases develop in the first place so that a cure can be developd. The cure development would thuys be a separate process.

If you dislike folding just because some people think it'll cure cancer then you're being irrational.
Peace.
 
The points update this morning seems a bit more varied (and thank you to all our new B3D PS3 folders). I see some 287, some 302, and some 350 point WUs
 
I dfind it pretty amazing PS3 f @ h performance crushes all other clients only a day or so after the client was released.

Now if only MS would make a client too so my 360 could addi ts power to the curing of cancer.. :cool:
..If I could stand the constant fan noise that is.
Peace.
 
I'm still confused about the claim that there's kinds of folding which CELL can't performe, makes no sense to me.

Reminds me of the ol' "SPEs can't do AI, physics and so on" claim.

Will someone please care to enlighten me? :idea:
 
I dfind it pretty amazing PS3 f @ h performance crushes all other clients only a day or so after the client was released.

Now if only MS would make a client too so my 360 could addi ts power to the curing of cancer.. :cool:
..If I could stand the constant fan noise that is.
Peace.
Not to sound like some biased fanboy but letting Xenon do folding might not be very good for the PR.
 
I'm still confused about the claim that there's kinds of folding which CELL can't performe, makes no sense to me.

Reminds me of the ol' "SPEs can't do AI, physics and so on" claim.

Will someone please care to enlighten me? :idea:


If the accusation holds any water at all, it would have to be over a system memory limitation not the Cell's computational ability.

Even on the desktop versions of FAH there is an advanced option to not accept the assignment of "extra large" Work Units. The memory foot print on some of them (but not the majority as far as I can gather) can be very large. So on lower end systems you can choose to reject them even on the PC version.
 
If the accusation holds any water at all, it would have to be over a system memory limitation not the Cell's computational ability.

Even on the desktop versions of FAH there is an advanced option to not accept the assignment of "extra large" Work Units. The memory foot print on some of them (but not the majority as far as I can gather) can be very large. So on lower end systems you can choose to reject them even on the PC version.
Thanks, that makes sense. :)
 
Not to sound like some biased fanboy but letting Xenon do folding might not be very good for the PR.
I think they could easily tap Xenos to do it, and perhaps use Xenon to do higher-precision work that Cell & GPUs don't do.
 
Back
Top