Okay, somebody had to be thinking about it. The iPhone runs a 650 mhz processor. It's got to be enough to be able to do a little bit of computing when it's in its charging cradle. Has anybody looked into a port?

This would be much more for the fun of it than for the actual contribution by the phone; I'm sure it wouldn't be a TON of science, but would be a fun experiment.

For starters, Apple doesn't yet allow 3rd party apps (except web ones running within Safari), and I don't know if they will ever allow it. Also, how much RAM does the phone have? Would it affect battery life considerably?

I think it would be a better idea to just control your computer's BOINC from the iPhone; now *that* would be the best BOINC Manager ever! Simple example: to attach, drag and drop projects from a list into the BOINC icon, with your finger

For battery life, I would assume to only have it run while the iPhone is charging. Of course it would use WAY too much battery to use just walking around



Very cool idea!

I do not know whether iPhone will be sold in China...

Wait a few years for Apple to debug it, improve battery life, make it Quad band UMTS (3G), and Quad band GSM before you buy it.

I know I am.

Yeah, don't hurry with the iPhone. Current owners are beta-testing it.

Jeesh... Now I have heard Europe is going to open another band to 3G so you will need a 5 band 3G 4 band GSM iPhone

I have an iPod touch. I can confirm anything that gets close to using all the CPU makes the iPod get hot enough to get me to worry. I don't think I would run BOINC on it for more than a few minutes... Not to mention the poor battery...


I have to correct you on that: both the iPhone and the iPod touch run at 412MHz.


I know somebody who ported a BitTorrent client to the iPhone. I challenged him to port BOINC next, still haven't heard back from him.

In any case, my previous suggestion about controlling your computer's BOINC from the iPhone/iPod touch still applies. I would do it if I knew any Objective-C and if I had time. (I already have the rest, like the compiler toolchain and the means to copy and run the app on the iPod)

I managed to compile BOINC for the iPhone (well, iPod touch; would run on iPhone too).

In case anyone still thought it would be a good idea: it's not.



(yes, I know I'm bumping an old thread)

Nice POV! Good job.

With benchmarks like that I suspect running CPDN or Orbit might not be advisable

OK, I have to ask, is that a photoshopped image, or is the iPod/iPhone really that slow?

Al.

It's a real image.

Well erm... the image is slightly fake. I really did run BOINC on the iPod, but not using MobileTerminal. I was accessing remotely with ssh. To take the screenshot, I ran "cat stdoutdae.txt" from MobileTerminal. Testing different things while running BOINC is easier to do on a full and real computer keyboard

So those are real benchmarks. However, it might be because the compiler I used just doesn't optimize too well for the ARM processor.

By the way, CPDN wouldn't be advisable due to the really little RAM available, not only for the processor

It's not a matter of optimizing for ARM. ARM was not designed for the ultimate performance, and neither are desktop processors, but for a good enough performance for a given power consumption. In the iPhone's case, a few Wh of battery energy, in a desktop's case, hundreds of Wh of outlet power. While ARM has just hundreds of thousands of transistors, a typical desktop processor has tens, hundreds of millions of transistors. What these numbers mean is that an ARM will not be so fast as a desktop processor.

HTH