we have been discussing this idea with Microsoft quite a bit over the past several weeks; I will keep everybody posted.

Looks like MS has choosen for F@h :(
http://www.dailytech.com/article.aspx?newsid=7238

---

Member of Dutch Power Cows

cureseeker, i've seen that elsewhere and it is just MS taking jobs at the PS3 and Sony.
They say they may look in to it for PR reasons if and when the need arises... ;-)

Nowhere have they actually said they are actively looking into it and producing a client... It just my cpu is as good as yours and we have the gpu oh and look a bigger userbase talk :)

---

Team mauisun.org

*cough* GPU *cough*

Any updates to this?

why is microsofts involvement needed? can't anyone write programs for it?

why is microsofts involvement needed? can't anyone write programs for it?

Yes, via XNA but I don't know it's license restrictions also it is an unknown platform from the Rosetta developers point of view.
Rom Walton had a look at XNA to see about porting BOINC to it, but there where some sort of complication (license ?), ease of doing it.

Other than using XNA I don't think any of the devs at boinc or rosetta know where to start.

So if you want you can port it across.
email rosetta admin or the mods about access to the rosetta source code.
pop to http://boinc.berkeley.edu/index.php to get the source and info from the boinc developers and get porting it. it is fully open source.

---

Team mauisun.org

I'm going to take a jab at this but my guess is Microsoft is waiting for new the chips in the 360 to come out before they release a folding program for it..?


What i don't understand is if its open source why hasn't anyone even bothered to make a folding program of any sort for the 360? PS3 has a ton for it besides just folding@home


Really sucks because my computer is not all that fast and buying a console is alot cheaper than buying a insane rig, but i have been folding every night espically when i go to bed, doing all i can at the moment :-)

I dont want to start a "console war" here but I really dont think the 360 can handle the heat that is generated when you are running a program like this. I work at Target and I have seen numerous people who are on their 2nd or 3rd 360 already, simply due to heat issues killing the machines. The PS3 is much better in this area and really has no heat issues. I have been running Folding@Home on my pretty much non stop since April 20 and already have over 130 WUs done.

Nothing against the 360 per se (ok the controller makes me nauseous :) ) but they need to fix the heat issues.

---

there's a new cooling system: http://www.dailytech.com/article.aspx?newsid=7667

---

As far as the PS3 being more powerful, I believe that while that's true in raw numbers, the xbox should be easier to program for, and I believe all 512MB of the RAM should be accessible for crunching whereas the PS3 only has 256MB available.

I need to touch this one. You're 100% right, but for 100% the wrong reasons.

The PS3 (cell) is far harder to code for, but it's because of the limitations of the SPU's, most notably memory. Each SPU has only 256K of local store, but it runs at effectively L1 cache speeds. So to get performance for Rosetta, we would need to take a WU, chop it into bite sized chunks which can run for a couple of ms on an SPU, while only using 80 to 100k of data.

That allows double buffering of data: DMA out N-1 results, DMA in N+1 data, all while N runs. At 100k per WU, that's 200K gone, which leaves 56K for code and "incidentals".

---

As far as the PS3 being more powerful, I believe that while that's true in raw numbers, the xbox should be easier to program for, and I believe all 512MB of the RAM should be accessible for crunching whereas the PS3 only has 256MB available.

I need to touch this one. You're 100% right, but for 100% the wrong reasons.

The PS3 (cell) is far harder to code for, but it's because of the limitations of the SPU's, most notably memory. Each SPU has only 256K of local store, but it runs at effectively L1 cache speeds. So to get performance for Rosetta, we would need to take a WU, chop it into bite sized chunks which can run for a couple of ms on an SPU, while only using 80 to 100k of data.

That allows double buffering of data: DMA out N-1 results, DMA in N+1 data, all while N runs. At 100k per WU, that's 200K gone, which leaves 56K for code and "incidentals".

surely the 256MB of RAM will be as much of a problem? If you have to have one thread per SPU then where are you going to store these threads while they're not in cache, seeing as the PS3 OS plus one rosetta thread will probably use all of that memory?

---

I dont want to start a "console war" here but I really dont think the 360 can handle the heat that is generated when you are running a program like this. I work at Target and I have seen numerous people who are on their 2nd or 3rd 360 already, simply due to heat issues killing the machines. The PS3 is much better in this area and really has no heat issues. I have been running Folding@Home on my pretty much non stop since April 20 and already have over 130 WUs done.

Nothing against the 360 per se (ok the controller makes me nauseous :) ) but they need to fix the heat issues.

2nd or 3rd ?

There's one guy on his TWELFTH 360

/me wouldn't touch em with a bargepole !!

---

I mentioned this a couple months ago, but as a recap:

There are about 4 million Xboxen in the world, each with 3 seperate 2ghz processors. Within the last few days, MS has released a software development tool (geared toward games obviously, but code is code) that allows groups to develop their own xbox apps.

I'm not a programmer. How difficult would it be to use those tools to port Boinc over to the Xbox?

12 million cpu's can't be wrong :)

http://arstechnica.com/news.ars/post/20060320-6420.html

-Ethan

What is up with Rosetta. Over a year and half, and NO PROGRESS on this. What a shame.

I mentioned this a couple months ago, but as a recap:

There are about 4 million Xboxen in the world, each with 3 seperate 2ghz processors. Within the last few days, MS has released a software development tool (geared toward games obviously, but code is code) that allows groups to develop their own xbox apps.

I'm not a programmer. How difficult would it be to use those tools to port Boinc over to the Xbox?

12 million cpu's can't be wrong :)

http://arstechnica.com/news.ars/post/20060320-6420.html

-Ethan

What is up with Rosetta. Over a year and half, and NO PROGRESS on this. What a shame.

Have you tried to help them ?
While it's nice to get an XBOX360 version, benchamrks of programs converted to just run on it use say XNA http://www.primatelabs.ca/blog/2007/08/xbox-360-performance-august-2007/ show the speed usage may not be so good.
It would take much more effort than it's probably worth unless Microsoft or someone else does it for them and gives them an easy program to compile with... Don't forget it would need to be compiled every time they update the rosetta client. It's a lot of testing for something that only has short term appeal at the moment. Same thing for the PS3.
It would make more sence currently to get a 64bit native client up and running, optimised to use 64bit if it can take any advantage (speed/accuracy/really really large memory tasks) Since we'll all have to end up using 64bit at some time int he future ;-)

You must remember that Rosetta (and the code) is much larger than just Rosetta@home. Also time and manpower may well be a problem.
Come on Seti@home, a simple program, is only just getting ported by individuals to GPU's. I see not client running on the PS3 or XBOX or iPhone ;-) yet and that a simple program.

---

Team mauisun.org

I mentioned this a couple months ago, but as a recap:

There are about 4 million Xboxen in the world, each with 3 seperate 2ghz processors. Within the last few days, MS has released a software development tool (geared toward games obviously, but code is code) that allows groups to develop their own xbox apps.

I'm not a programmer. How difficult would it be to use those tools to port Boinc over to the Xbox?

12 million cpu's can't be wrong :)

http://arstechnica.com/news.ars/post/20060320-6420.html

-Ethan

What is up with Rosetta. Over a year and half, and NO PROGRESS on this. What a shame.

indeed it is a shame, I can't see why they haven't looked into it completely, I even made a petition XD

You make some good points, Fluffychicken.

I do think that porting to GPUs (if it's possible with the kind of science that Rosetta does) would be the best thing to do right now, as long as it isn't too specific to a single GPU card and the code will be able to be used for a long time on progressively faster cards.

That could give quite a big TFLOPS boost to the project and allow Dr. Baker & co run more projects in parallel and get better results on them too (lets kick butt at CASP 8!).

---

i read somewere that in order to run boinc on your ps3 you need to install linux on it, so if you could tho the same with your xbox, i thought it had a harddisk or something like that, you can just run linux and boinc for linux, so no new programming needed.

I guess "David Baker" has failed with Microsoft.

And judging from the lack of responses to my questions, doesn't even want to try with Sony.

I guess "David Baker" has failed with Microsoft.

i read somewere that in order to run boinc on your ps3 you need to install linux on it, so if you could tho the same with your xbox, i thought it had a harddisk or something like that, you can just run linux and boinc for linux, so no new programming needed.

(i know it's an old post) installing linux on the xbox isn't a trivial task due to the hypervisor and other security measures - free60.org are working on it but it's not possible ATM (in any worthwhile context) without MS's approval...

---