Orgil, I don't know how to say this without sounding like a Hallmark card but...

I saw this, and I thought of you!:D

DailyTech - Rise of the Machines - Mobile Devices to Outnumber Humans by 2016

But synbio is the answer. Better watch those related scientific broadcasts which is pretty cool.

---

Mandtugai!

put it simple a phone can not crunch you prove me wrong have some one write an app for crunching on any smart phone and show me the rac.

http://setiathome.berkeley.edu/show_host_detail.php?hostid=6383032
It's a HTC Desire (the non-HD version). Too unstable to be useful yet, but hey it's a proof of concept anyways.

Have a look in the BOINC Wiki for more crazy phone stuff...

For those who don't like to read, here's the boiled down version:

• It is slow, SLOW
  
• Heat is not much of an issue (wrt overheating)
  
• Battery on a day-to-day basis is not an issue
  
• Long term there may be some detrimental effects on battery
  
• It is possible to go a factor of around 10 faster than this on the same hardware
  

Oh, and while talking about things that you can't do: you also you can't do this on my old HTC hero =)

all 3 WU's failed with result overflow. Nice try. Thank You come again.

The one wingman that has returned results didn't match yours. I'm willing to bet you are going to get "0" credit for your efforts
http://setiathome.berkeley.edu/result.php?resultid=2295085922

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

Nice try. Thank You come again.

Nice attitude btw...

These aren't the only results returned by the phone (it does have credit for some earlier runs). The most recent workunits are experiments with a newer version of libfftw3 that supports the GPU on the phone. It seems the old WUs are no longer in the list.
Interestingly enough, the new code seems to pick up more tripplets and pulses than regular clients - so it's basically the same stuff + some more.
You should, however, be able to get an idea of the WU runtime compared to a normal client with this.

Nice try. Thank You come again.

Nice attitude btw...

These aren't the only results returned by the phone (it does have credit for some earlier runs). The most recent workunits are experiments with a newer version of libfftw3 that supports the GPU on the phone.
Interestingly enough, the new code seems to pick up more tripplets and pulses than regular clients - so it's basically the same stuff + some more.
You should, however, be able to get an idea of the WU runtime compared to a normal client with this.

Ty and again no you can't get any idea since the WU's ran into a 30 result wall. There is no way of knowing how long they would have ran had they not gotten 30+ results.

Your optimism is noted and repelled by my scepticism.

the speed differnce alone should tell you something went wrong on your socalled smart phone.

your wingman
Measured floating point speed 3014.38 million ops/sec
Measured integer speed 9181.49 million ops/sec

the smart phone

Measured floating point speed 386.55 million ops/sec
Measured integer speed 1658.23 million ops/sec

Notice anything diffent in those speeds.

thats anywhere from a 5-10X difference

(For giggles here's my AMD 3000+ numbers

Measured floating point speed 2069.1 million ops/sec
Measured integer speed 3438.31 million ops/sec
thats a 2.1Ghz CPU Fiugre your's runs about 1/3 of that speed)


which means you'd have spent a minimum of 82,325 seconds completing that task. That calculates out to almost 23 hours. Minimum worst case scenario is about 46 hours. still not impressive. Then lets figure in that your wingman isn't using an optimized app which could have shortened his working time by almost 50%

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

Your optimism is noted and repelled by my scepticism.

Mhm.

[Edit:] The numbers you are quoting in the edit of your post are based on measurements from the BOINC client which was not built with the same compile flags as the actual science application.
Actual tests suggests that without the phone GPU it takes around 2-5 days to complete a WU that a relatively new laptop takes less than 8 hours to complete, or a computer GPU takes less than half an hour to complete.

[Edit 2:] And I think you may have misunderstood my point - I'm saying that it is really slow, not the opposite.

put it simple a phone can not crunch you prove me wrong have some one write an app for crunching on any smart phone and show me the rac.

http://setiathome.berkeley.edu/show_host_detail.php?hostid=6383032
It's a HTC Desire (the non-HD version). Too unstable to be useful yet, but hey it's a proof of concept anyways.

Have a look in the BOINC Wiki for more crazy phone stuff...

For those who don't like to read, here's the boiled down version:

• It is slow, SLOW
  
• Heat is not much of an issue (wrt overheating)
  
• Battery on a day-to-day basis is not an issue
  
• Long term there may be some detrimental effects on battery
  
• It is possible to go a factor of around 10 faster than this on the same hardware
  

Oh good it is already in the works. So when I get a new phone I can root my current phone and get to playing with it.

With the CPU clocking tools out there it should not be to hard to find a speed for it to run at where things don't get to hot.

As I would no longer be using it as my cell phone I would probably take the battery out and just run it off USB power. Since leaving it connected with the battery in would just end up killing the battery anyway.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

it should not be to hard to find a speed for it to run at where things don't get to hot.

They generally don't get particularly hot at all while only using the CPU. Most of the phones are actually underclocked from the factory and will easily overclock by 10% without getting any hotter (no core voltage increase required).

Worst case is while charging and crunching at the same time.

it should not be to hard to find a speed for it to run at where things don't get to hot.

They generally don't get particularly hot at all while only using the CPU. Most of the phones are actually underclocked from the factory and will easily overclock by 10% without getting any hotter (no core voltage increase required).

Worst case is while charging and crunching at the same time.

Are you able to monitor the CPU temp in your device? Mine doesn't seem to have, or at leas report that it has, a CPU thermal monitor. However, I do get to monitor the battery temp. I expect most of the heat build up that people see is from the battery charging or discharging.

Mine tends to idle at 245.76MHz when not plugged in and 480MHz when it is connected to a half amp USB port. My CPU is set for a max of 600MHz, but in other devices the same CPU it was set for 800MHz. So one of those overclocking apps should easily let me run 800MHz or more. A guy I work with run his nook color at 1.2GHz instead of the stock 800MHz

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Nope, I too have been using the battery temperature monitor and an external thermometer. I didn't check for in-core sensors though, it may have them.

put it simple a phone can not crunch you prove me wrong have some one write an app for crunching on any smart phone and show me the rac.

http://setiathome.berkeley.edu/show_host_detail.php?hostid=6383032
It's a HTC Desire (the non-HD version). Too unstable to be useful yet, but hey it's a proof of concept anyways.

Have a look in the BOINC Wiki for more crazy phone stuff...

For those who don't like to read, here's the boiled down version:

• It is slow, SLOW
  
• Heat is not much of an issue (wrt overheating)
  
• Battery on a day-to-day basis is not an issue
  
• Long term there may be some detrimental effects on battery
  
• It is possible to go a factor of around 10 faster than this on the same hardware
  

Here's what found

Name 07oc11ag.29450.18885.13.10.38.vlar_2
Workunit 922942129
Created 2 Feb 2012 | 7:33:28 UTC
Sent 2 Feb 2012 | 10:13:17 UTC
Received 11 Feb 2012 | 19:24:27 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x0)
Computer ID 5485689
Report deadline 21 Mar 2012 | 4:08:21 UTC
Run time 16,464.85
CPU time 14,798.32
Validate state Checked, but no consensus yet
Credit 0.00
Application version SETI@home Enhanced v6.03
Stderr output

<core_client_version>6.12.34</core_client_version>
<![CDATA[
<stderr_txt>
setiathome_enhanced 6.02 DevC++/MinGW
libboinc: 6.3.6

Work Unit Info:
...............
WU true angle range is : 0.066636
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_vGetPowerSpectrumUnrolled 0.00007 0.00000
sse3_ChirpData_ak 0.00996 0.00000
v_vTranspose4ntw 0.00638 0.00000
BH SSE folding 0.00158 0.00000

Flopcounter: 49489131273635.562000

Spike count: 3
Pulse count: 11
Triplet count: 1
Gaussian count: 0
called boinc_finish

</stderr_txt>
]]>


and Here

Owner Janus
Created 28 Jan 2012 | 16:46:29 UTC
Total credit 229
Average credit 13.14
Cross project credit
CPU type
Number of processors 1
Coprocessors ---
Operating System Linux
2.6.37.6-cyanogenmod-g37556d9
BOINC version 7.0.11
Memory 405.53 MB
Cache -0 KB
Measured floating point speed 386.55 million ops/sec
Measured integer speed 1658.23 million ops/sec
Average upload rate 8.08 KB/sec
Average download rate 294.74 KB/sec
Average turnaround time 1.77 days
Application details Show
Tasks 4
Number of times client has contacted server 61
Last contact 16 Feb 2012

looks like not much to me except he can play a game on his phone so far no cruncher and no rac.

---

[snip]
sse3_ChirpData_ak 0.00996 0.00000
[snip]

Looks like the wrong WU link to be honest - that's from a normal computer (the ARMv7a chip has VFP3 not SSE). For some reason the first workunits done on the phone (those that didn't fail) are no longer listed on the website.

I did try out a few Einstein apps as well earlier, they took a lot longer to run on the phone (and more memory). They all started becoming increasingly unstable about two thirds through the workunits (after 5-7 days) and their results were clearly not valid when compared to the same unit computed on a computer. They're assuming SSE2-kinda-floats and the floating point units in the phones are not entirely compatible with that - which is probably why it fails.
SETI on the other hand completes the reference workunit fine and fits well with phone memory. If you disable some of the optimizations it will even use less than 64MB of RAM but will run slower (even slower!).

put it simple a phone can not crunch you prove me wrong have some one write an app for crunching on any smart phone and show me the rac.

http://setiathome.berkeley.edu/show_host_detail.php?hostid=6383032
It's a HTC Desire (the non-HD version). Too unstable to be useful yet, but hey it's a proof of concept anyways.

Have a look in the BOINC Wiki for more crazy phone stuff...

For those who don't like to read, here's the boiled down version:

• It is slow, SLOW
  
• Heat is not much of an issue (wrt overheating)
  
• Battery on a day-to-day basis is not an issue
  
• Long term there may be some detrimental effects on battery
  
• It is possible to go a factor of around 10 faster than this on the same hardware
  

Hey Janus the biggest congrats to you!! If possible could you try Boinc on Samsung Galaxy S2 which that phone has dual core cpu that usually 150% higher benchmark score than older 1ghz cpu phones. Maybe asking your friends or any close contacts. That should be even more interesting trial. And dual core phones should have even more advanced chipset to handle crunching applications normally as HD games.

Your finding of less heating in phone crunching should be noted as DISCOVERY OF THE MONTH in Boinc world!!

Also LG and Samsung are launching new quad core phones almost now which according to their claim that is 5x higher cpu power of phones than dual core phones at same time equal or less power consuming tech solution to dual cores as their claim.

---

Mandtugai!

I guess the pc desktop is safe from the hordes of phone crunchers.
I have yet to see a smart phone cruncher app that can be installed and produce a rac. Point me to one if one exist.

---

The new quad cores are promising stuff after all:

LG x3

Looks like quad cpu really is several times higher than a dual core.

While little older benchmark for dual cores were like this:

for SGS2

---

Mandtugai!

What EVER.........proof of concept.......perhaps.

---

"Time is simply the mechanism that keeps everything from happening all at once."

The grand assumption with smart phones is that suddenly CPU and GPU technology will suddenly stop advancing. That's hardly evident from current CPU types. Sandybridge and Bulldozer alike. GPU's also are advancing faster than ever before. To say that a smartphone is suddenly going to leapfrog over these technologies is silly. I wouldn't want a smartphone and only use my recently purchased laptop for traveling. The recent inundation of tablets and ipads are pretty much a fad. It's new and every status seeker is out to have the latest greatest gadget regardless of how outrageously its priced or how limited its capability. It looks flash so the little child must have that flashy item. Bah

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

The grand assumption with smart phones is that suddenly CPU and GPU technology will suddenly stop advancing. That's hardly evident from current CPU types. Sandybridge and Bulldozer alike. GPU's also are advancing faster than ever before. To say that a smartphone is suddenly going to leapfrog over these technologies is silly. I wouldn't want a smartphone and only use my recently purchased laptop for traveling. The recent inundation of tablets and ipads are pretty much a fad. It's new and every status seeker is out to have the latest greatest gadget regardless of how outrageously its priced or how limited its capability. It looks flash so the little child must have that flashy item. Bah

Thank you skildude, Thats the point that I havent been able to get across in my posts.

At the moment smart phones are weed whackers, And they do a good job at what they were designed to do. My desk top is a chain saw and Boinc is the tree. Me I will cut the tree down with a chainsaw not a weed whacker.

---

[/quote]

Old James

The grand assumption with smart phones is that suddenly CPU and GPU technology will suddenly stop advancing. That's hardly evident from current CPU types. Sandybridge and Bulldozer alike. GPU's also are advancing faster than ever before. To say that a smartphone is suddenly going to leapfrog over these technologies is silly. I wouldn't want a smartphone and only use my recently purchased laptop for traveling. The recent inundation of tablets and ipads are pretty much a fad. It's new and every status seeker is out to have the latest greatest gadget regardless of how outrageously its priced or how limited its capability. It looks flash so the little child must have that flashy item. Bah

Thank you skildude, Thats the point that I havent been able to get across in my posts.

At the moment smart phones are weed whackers, And they do a good job at what they were designed to do. My desk top is a chain saw and Boinc is the tree. Me I will cut the tree down with a chainsaw not a weed whacker.

In that kind of analogy I would say this:
Mobile Devices=Weed whacker
CPUs=Push mower
GPUs=Riding mower
Work=Half acre lot of grass

Granted no one wants to do that much work with a weed whacker, but if someone wants to spend the time. Then I don't see why they shouldn't. It is their time to spend.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[