Each new Xbox 360 has 3 multi-ghz CPUs. . . and with the way it's being advertised, the Xbox will be on a lot of the time not utilizing all of them.

Just about all of these machines are on some sort of network connection. . .

So, a couple million Xboxen (eventually), three cpus each, and it runs some sort of Microsoft OS.

It shouldn't be hard for MS to build a boinc client that folks could opt in.

Bill Gates is a huge contributor to causes such as AIDS and malaria research, sound familiar?

It's great PR for Microsoft, users who find 'significant results' can still be tracked by their Live account, and obviously it would be a huge benefit for the project.

Anyone have contacts in Redmond?

-Ethan

---

If anything it would be more like a linux client that would run on them,I can't see Microsoft doing this at all.

there has already been a linux operating system that can work on modded Xbox clients so maybe one will be released for Xbox 360

---

Join us in Chat (see the forum) Click the Sig


Join UBT

Linux on XBox 360; it won't be easy

---

Linux will be hard, MS learned their lesson with Xbox 1.

I'm not suggesting people try to make a port. . I'm suggesting a fully tested client designed to run on the Xbox 360.

It could be as easy to install as it is to download a movie trailer or buy a game. . the infrastructure is already in place with Xbox-live, all that is needed is the support from MS. . .

-E

---

Each new Xbox 360 has 3 multi-ghz CPUs. . . and with the way it's being advertised, the Xbox will be on a lot of the time not utilizing all of them.

Your best bet would be to get in touch with a 3rd party developer, and put the idea to them. If they can present it to Microsoft as a "fait accompli" it'd have a far higher likelyhood of seeing the light of day.

While you're at it, the Cell processor in the PS/3 should also be able to do this. That'd be another avenue to investigate.

If you'se guys are interesting in following up with this, I'd suggest putting it on the back burner till March 2006. I'll probably take a day or two at the Game Developers Conference this year, so I can poke round and see if there's any interest at all.

The idea of using the Xbox 360 or PS/3 as a low end farm machine has a certain charm - they're dirt cheap when you look at the CPU horsepower / dollar ratio, but then any console is.

---

Anyone have contacts in Redmond?

Why not try Scobleizer or Major Nelson with direct emails? Heck, you are all in Seattle.

---

---

So, a couple million Xboxen (eventually), three cpus each, and it runs some sort of Microsoft OS

An honorable idea, but it won't happen IMHO. Anything that could destabilize a platform already getting bad press for its frequent crashes just won't get anywhere. It's the same reasonning that big companies have for not installing DC projects on their thounsands of idle computers: it's just not worth the risk - even if the risk is nil. And from experience the risk isn't really nil - overheating killed one of my boxes - and it was directly boinc related.

Another thing about the heat - the XBox 360 overheats very easily - something like boinc would kill it in no time.

Your only hope is that linux be ported to the 360 - the other posts go over that in details.

---

Team CFVault.com
http://www.cfvault.com

I had another thought last night. . the xbox would be great because it has a distribution system and 3 fast cpus. . but there are 'only' going to be a few million xboxen, there are already 100+ million pc's running windows. What if MS included Boinc functionality in Vista or as an optional update via windows update? Both would be opt in, just as with the xbox.

Any client, xbox or pc, would have to make it clear what the user is accepting. Their machine will likely consume more power and require sufficient ventilation, but in return, your machine will be working on a cutting edge research project with great potential.

Thanks for the comments/ideas/suggestions so far. . . I had thought of cell, and as you point out, it is much more specialized than the xbox core. Xbox/PCs are the 'low hanging fruit' in my opinion. . things like cell and gpu's are a big resource that may be worth looking at down the road.

---

Thanks for the comments/ideas/suggestions so far. . . I had thought of cell, and as you point out, it is much more specialized than the xbox core. Xbox/PCs are the 'low hanging fruit' in my opinion. . things like cell and gpu's are a big resource that may be worth looking at down the road.

I'm sure you are familiar with what F@D is planning to do with the X1800 GPUs - yes this is indeed very exciting!

---

Team CFVault.com
http://www.cfvault.com

Thanks for the comments/ideas/suggestions so far. . . I had thought of cell, and as you point out, it is much more specialized than the xbox core. Xbox/PCs are the 'low hanging fruit' in my opinion. . things like cell and gpu's are a big resource that may be worth looking at down the road.

Getting to the SPU's on the cell, which is where you have to go to get those large GFLOPS (and maybe TFLOPS) numbers, will be problematic for Boinc applications like Rosetta et al. No arguments there, none at all.

However, the PPC core is a very conventional CPU with a half way decent amount of memory, and trucking along at a reasonable rate of knots. THAT would be a no-brainer to code for, espectially since the PPC is one of the cpu's that you can find in the MAC, hence we already have a CPU specific port. And with Linux looking hopeful, combined with the fact that the current MAC OS is basically Unix in drag, it also falls into the "low hanging fruit" category.

---