Phones are just now becoming powerful enough to run some seti software if it was geared towards them. Quad core processor phones will be common place after this coming summer. You would just have to solve the battery drainage issue.. if even a small fraction of cell phone users installed the app im sure there would be insane results. I for one would for sure install it on my phone.

There's no way to solve battery drainage when your running something that is meant to process as much as possible.

Even though phones may be soon powerful enough to run some units for seti or other Boinc projects, it would only be worth doing it when you were plugged in.

I'm not against a future app for android that crunches seti, and you are not the first person to bring up the idea. But there are drawbacks to be considered. And quad core phones will be out this summer, but "commonplace", that wont happen for another 18 months to 2 years.

I do see some value in this idea, especially for projects like "quake catcher network" that use vibration sensors in laptops. Smartphones have many sensors that cant be found even on a laptop. perhaps a Boinc app for smartphones could be used for a project that will specifically take advantage of smartphone sensors without being processor intensive..

---

#resist

The problem is, cell phones are built with the idea of "just enough" computing power for what they do. Sure, they may have a quad core processor, but that quad core isn't the same as your quad core in your PC.

In fact, a cell phone's quad core, while lacking a strong floating point math coprocessor, would be as powerful as a Pentium II CPU. Since most of SETI@Home math calculations are handled by the floating point unit, this would only be somewhat beneficial.


Then there's the heat issue. Any processor working at max operation, such as how SETI tends to run CPUs, put off a lot of heat as a byproduct. Cell phones were not designed to dissipate the excess heat from 100% CPU utilization over long periods of time (notice how hot they get when you use the phone for a long time?).


Lastly, there is no real way around the battery draining issue. If you're actively using a portion of the phone, even if you were to limit the CPU power to 20% per core, you're still using the battery, which will add up considerably.

It has been suggested that the SETI processing should be limited to only run while the battery is on the charger, but then you're right back at the heat issue. How upset would you be if you woke up in the morning to find that your cell phone is completely cold and won't turn on because it burned out somewhere in the middle of the night?



I seriously doubt that any cell phone carrier would allow you to run SETI@Home and still get a warranty from them. No one would want to touch the risk of users burning out their cell phones and wanting to have it replaced for next to nothing. They already have a clause for water damage, it won't be that hard to add a "no distributed computing" clause.

Warranty? What warranty? I root my phones as soon as I can. :-)

And ruling out shared computing just because it can be hot is not really a good idea. Like I said someone could design a Boinc project that takes advantage of smartphone sensors, without being processor intensive. I think that could be a worthwhile idea, as it could serve a real purpose without being of any real strain on a phone.

---

#resist

LOL Well, not everyone roots their phones. I'm a tech and I refuse to root my phone.


Sure, it is possible to make a project geared specifically for phones, and with the way everything is going over to mobilization, it might be the only chance for distributed computing to have a chance of surviving.

Unfortunately, I still stand by my opinion that mobile devices make a trade-off between power and battery life / heat dissipation, and distributed computing offsets that trade-off greatly beyond the device designer's specifications.

Someone is working on BOINC for Android. iOS (Apple) licensing restrictions are intentionally impossible for open source projects to meet. Therefore there will not be an iOS application for the foreseeable future.

[edit]

I am still unconvinced that an application for a handheld device is worth the effort. They do not have floating point processors. They have battery restrictions. They have no cooling.

I remember an application we wrote that had a bug that would lock the device in the on state. Battery life was reduced from several days to a couple of hours.

---

BOINC WIKI

I'll mention it again. Since everyone seems to think a boinc project has to be processor intensive... :-)

Indeed, an Android app wont do much for seti@home. But...

Someone could design a boinc project that is not centered around processing, but instead focuses on data collection using the very powerful suite of standard sensors on Android (and iOS) phones.

I'm sure there could be countless uses for data collected from some of these sensors.

Quake catcher network for example uses the vibration sensors in laptops as a cumulative seismic data collector. I doubt that application relies on much CPU power at all..

So I, see the potential for many powerful uses for boinc here (on smartphones) that do not need to use the processor much at all.

There could even be a project with two halves, one half used by people on smartphone boinc to collect data, and the other half ran on users machines' boinc to crunch whatever data.

If I had time, and money, and connections I could come up with some great ideas and implementation for such projects.

---

#resist

While its a great idea that you're talking about a general BOINC app that does other things, the OP and many other people are asking about a SETI@Home app.


Its not up to BOINC, nor the BOINC framework to come up with a project idea and then publish an app for it. Anyone is free to make their own project and then develop an app for whatever platform they wish to.


We're not arguing that cell phones are completely useless for distributed computing... we're arguing that cell phones are not great for processor intensive projects like SETI@Home.

While its a great idea that you're talking about a general BOINC app that does other things, the OP and many other people are asking about a SETI@Home app.


Its not up to BOINC, nor the BOINC framework to come up with a project idea and then publish an app for it. Anyone is free to make their own project and then develop an app for whatever platform they wish to.


We're not arguing that cell phones are completely useless for distributed computing... we're arguing that cell phones are not great for processor intensive projects like SETI@Home.

Agreed.

Indeed, an Android app wont do much for seti@home. But...

---

#resist

Its not up to BOINC, nor the BOINC framework to come up with a project idea and then publish an app for it. Anyone is free to make their own project and then develop an app for whatever platform they wish to.

And no it's not BOINCs job to publish an app for a specific project, but publishing BOINC to Android is a different thing entirely, and something that should be considered.

So I guess I should design a project then. If I could just find the time, and oh yea, learn some real programming skills. ;-)

---

#resist

Well... BOINC is open source so anyone could port it to Andriod or iOS. ;-D

Emm, have any of you guys tried using shell command and running a linux/uniux. Version of boinc that way?
I also see on play market linux installers.

---

Remember that you need hardware support as well. Just because Linux runs on/is Android, and BOINC runs on Linux, doesn't mean that BOINC will run on Android on an ARM processor.

Every app has to be re-compiled for a different hardware architecture before it will run on it natively.


However, according to this thread, it looks like someone indeed re-compiled BOINC to run on an ARM processor and thus any Android smartphone.

However, according to this thread, it looks like someone indeed re-compiled BOINC to run on an ARM processor and thus any Android smartphone.

Or see http://boinc.berkeley.edu/dev/forum_thread.php?id=7414.

---