Smartphone crunching

Message boards : Number crunching : Smartphone crunching
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 6 · 7 · 8 · 9 · 10 · 11 · 12 . . . 13 · Next

AuthorMessage
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1200243 - Posted: 27 Feb 2012, 3:56:35 UTC - in response to Message 1200161.  
Last modified: 27 Feb 2012, 3:59:16 UTC

considering the additional cost for the bandwidth its just cheaper and simpler to have a home broadband connection that doesn't limit your usage

also the 3960 has a much higher Mflops than you think. In other words you read the wrong part of the test. What in Gods name would make you think a handle held device had as much or more crunching power as a 6 core CPU



The important things you missed are the 1st and 3rd test on that list. Its hard to read but the numbers look a lot better than you assumed. I think its odd that the amd FX-8150 out ran the 3960X on the FPU. Considering that the 8150 only has 4 FPU's and shares them.


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1200243 · Report as offensive
Profile AlphaLaser
Volunteer tester

Send message
Joined: 6 Jul 03
Posts: 262
Credit: 4,430,487
RAC: 0
United States
Message 1200260 - Posted: 27 Feb 2012, 5:27:38 UTC - in response to Message 1200243.  


also the 3960 has a much higher Mflops than you think. In other words you read the wrong part of the test. What in Gods name would make you think a handle held device had as much or more crunching power as a 6 core CPU


Nowhere do I even suggest that, in fact I was hoping to show the opposite. Despite the recent innovations in smartphone SoC's there's still a large gap between them and desktop CPUs. What is interesting though is that those chips are quickly advancing and may eventually become powerful enough to be relevant for volunteer computing. Maybe not SETI at first as other projects may be less FP-intensive or have smaller, more manageable WU lengths.

I think its odd that the amd FX-8150 out ran the 3960X on the FPU. Considering that the 8150 only has 4 FPU's and shares them.


That is weird. In most benchmarks I've seen Bulldozer suffer due to the FPU contention even compared to Phenom II. For example:


ID: 1200260 · Report as offensive
hbomber
Volunteer tester

Send message
Joined: 2 May 01
Posts: 437
Credit: 50,852,854
RAC: 0
Bulgaria
Message 1200309 - Posted: 27 Feb 2012, 11:37:24 UTC - in response to Message 1200123.  


Core i7-3960X (Stock, HT On) - 64.9275 GFLOPS
[url=http://www.xbitlabs.com/articles/cpu/display/core-i7-3960x-3930k_12.html]Core i7-3960X (4.5 GHz, HT On) - 133.9831 GFLOPS

I guess in second case is used Linpack with AVX, this explains the difference.

ID: 1200309 · Report as offensive
Profile shizaru
Volunteer tester
Avatar

Send message
Joined: 14 Jun 04
Posts: 1130
Credit: 1,967,904
RAC: 0
Greece
Message 1202550 - Posted: 5 Mar 2012, 3:34:12 UTC

I struck gold! You guys won't believe the date on this thread:

http://setiathome.berkeley.edu/forum_thread.php?id=32889

It's been almost 6yrs and it looks like we still need a couple more, before crunching on a smartphone...
ID: 1202550 · Report as offensive
Profile Link
Avatar

Send message
Joined: 18 Sep 03
Posts: 834
Credit: 1,807,369
RAC: 0
Germany
Message 1232427 - Posted: 15 May 2012, 10:28:39 UTC - in response to Message 1202550.  

It's been almost 6yrs and it looks like we still need a couple more, before crunching on a smartphone...

Well, eventually not few years more... just found that on Milkyway forums: Boinc Client for Android.
And this smartphone is crunching Milkyway WUs at a performance close to that of my AthlonXP 2000+. And it generates valid results, so it's not just some experimental thing, but actually something that can be used by anyone who wants to.
ID: 1232427 · Report as offensive
Profile shizaru
Volunteer tester
Avatar

Send message
Joined: 14 Jun 04
Posts: 1130
Credit: 1,967,904
RAC: 0
Greece
Message 1232445 - Posted: 15 May 2012, 12:03:42 UTC - in response to Message 1232427.  
Last modified: 15 May 2012, 12:20:31 UTC

That's awesome!

Average processing rate 1.1347731300531
I wonder what APR uses as a reference... What is APR=1?
(you'd think it would be a cobblestone per day, but seems to be almost exactly half of that)

When I said "two years", I was thinking more about Cuda & OpenCL widely available to the smartphone public. Of course, I should be writng these things down... instead of keeping them in my head:)

EDIT: It's a Galaxy SII
we are only one step from installing Boinc on fridges, washing mashines, cars... :-D
ID: 1232445 · Report as offensive
Andy Williams
Volunteer tester
Avatar

Send message
Joined: 11 May 01
Posts: 187
Credit: 112,464,820
RAC: 0
United States
Message 1232446 - Posted: 15 May 2012, 12:11:29 UTC
Last modified: 15 May 2012, 12:12:42 UTC

The numbers I find wildly optimistic in this thread are the estimates of the number of smartphone users who would participate. Does anyone know the number of active users SETI@Home has now? World-wide, as of today, 161,483 who have returned a result in the last 30 days.

Estimates of 50,000,000 or even 500,000 or perhaps even 50,000 smartphone users who would put up with the inconvenience and possible battery-life-shortening consequences seem unrealistic to me.
--
Classic 82353 WU / 400979 h
ID: 1232446 · Report as offensive
Profile James Sotherden
Avatar

Send message
Joined: 16 May 99
Posts: 10436
Credit: 110,373,059
RAC: 54
United States
Message 1232455 - Posted: 15 May 2012, 12:52:22 UTC

I just bought a smartphone 3 weeks ago. I can feel the thing warming up just from doing a google search, Or Playing solataire. No way would I crunch on it.

My data plan is for 4GB a month. Im am wondering what crucnhing 24/7 would do to that.
[/quote]

Old James
ID: 1232455 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14653
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1232458 - Posted: 15 May 2012, 12:54:38 UTC - in response to Message 1232445.  

That's awesome!

Average processing rate 1.1347731300531
I wonder what APR uses as a reference... What is APR=1?
(you'd think it would be a cobblestone per day, but seems to be almost exactly half of that)

APR is measured in gigaflops (billions of floating point operations per second). I'm slightly surprised that a smartphone has such a strong FPU, but that's what BOINC thinks as well...

EDIT: It's a Galaxy SII

CPU type: ARMv7 Processor rev 1 (v7l) @1200MHz
Measured floating point speed: 996.37 million ops/sec
Measured integer speed: 1972.43 million ops/sec
ID: 1232458 · Report as offensive
Profile Link
Avatar

Send message
Joined: 18 Sep 03
Posts: 834
Credit: 1,807,369
RAC: 0
Germany
Message 1232460 - Posted: 15 May 2012, 13:00:58 UTC - in response to Message 1232445.  

Average processing rate 1.1347731300531
I wonder what APR uses as a reference... What is APR=1?
(you'd think it would be a cobblestone per day, but seems to be almost exactly half of that)

No idea, my HD3850 could make at Milkyway about 34,000 per day if running 24/7 and has an APR of 36.47, so one could think here it would be (at least for the MW separation app) 1000 Credits per day, however that does not add up for my Pentium M laptop with an APR of 0.64 and something around 300 credits/day.

The smartphone I posted above can make according to the numbers in his task list 310-330 credits per CPU-day (using just 1 of his two cores?), so it seems to be even a little faster than my Laptop... and that makes it not-so-little faster than my AthlonXP.

Note to myself: eventually it's time for some upgrades... ;)
ID: 1232460 · Report as offensive
Profile James Sotherden
Avatar

Send message
Joined: 16 May 99
Posts: 10436
Credit: 110,373,059
RAC: 54
United States
Message 1232462 - Posted: 15 May 2012, 13:05:40 UTC

I have to say that I am impressed. But what is the comparison from milkyway to Seti@home as far as working the CPU power wise?
[/quote]

Old James
ID: 1232462 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1232473 - Posted: 15 May 2012, 13:41:06 UTC - in response to Message 1232455.  

I just bought a smartphone 3 weeks ago. I can feel the thing warming up just from doing a google search, Or Playing solataire. No way would I crunch on it.

My data plan is for 4GB a month. Im am wondering what crucnhing 24/7 would do to that.

My 30 or so machines at work use about 3.5GB per month for SETI@Home. Which includes all traffic to/from berkeley.edu & not just the tasks transferred.

If your smartphone can do work fast enough to transfer more data than that I should like one to replace those 30 machines I'm using. :D

With many review sites comparing various smartphones & tablets I would hope that we might be able to use that as a scale to compare S@H performance.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1232473 · Report as offensive
Profile James Sotherden
Avatar

Send message
Joined: 16 May 99
Posts: 10436
Credit: 110,373,059
RAC: 54
United States
Message 1232531 - Posted: 15 May 2012, 15:44:56 UTC - in response to Message 1232473.  

I just bought a smartphone 3 weeks ago. I can feel the thing warming up just from doing a google search, Or Playing solataire. No way would I crunch on it.

My data plan is for 4GB a month. Im am wondering what crucnhing 24/7 would do to that.

My 30 or so machines at work use about 3.5GB per month for SETI@Home. Which includes all traffic to/from berkeley.edu & not just the tasks transferred.

If your smartphone can do work fast enough to transfer more data than that I should like one to replace those 30 machines I'm using. :D

With many review sites comparing various smartphones & tablets I would hope that we might be able to use that as a scale to compare S@H performance.


Thank you for that info Hal, I would have guessed you would have used a lot more. Not that I will start crunching on my smartphone but there is one obstacle out of the way.
[/quote]

Old James
ID: 1232531 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1232534 - Posted: 15 May 2012, 15:53:44 UTC - in response to Message 1232531.  

I just bought a smartphone 3 weeks ago. I can feel the thing warming up just from doing a google search, Or Playing solataire. No way would I crunch on it.

My data plan is for 4GB a month. Im am wondering what crucnhing 24/7 would do to that.

My 30 or so machines at work use about 3.5GB per month for SETI@Home. Which includes all traffic to/from berkeley.edu & not just the tasks transferred.

If your smartphone can do work fast enough to transfer more data than that I should like one to replace those 30 machines I'm using. :D

With many review sites comparing various smartphones & tablets I would hope that we might be able to use that as a scale to compare S@H performance.


Thank you for that info Hal, I would have guessed you would have used a lot more. Not that I will start crunching on my smartphone but there is one obstacle out of the way.

With the download data being about 370KB and the return data being about 30KB you could use 400KB per tasks to calculate your usage. So if you did 1 tasks a day that would be about 12MB per month. Then you can take 12MB * tasks per day. So 100 per day would be about 1.2GB or so for a rough number.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1232534 · Report as offensive
Profile Julie
Avatar

Send message
Joined: 15 May 12
Posts: 279
Credit: 126,042
RAC: 0
United States
Message 1232654 - Posted: 15 May 2012, 23:36:05 UTC - in response to Message 1232534.  
Last modified: 16 May 2012, 0:07:09 UTC

(Hello, I looked around for an introduction thread and didna see one. On that note... hi :)

Reading through this thread was fun.
I have a Galaxy S and I tried to envision it crunching packets and all I could visualize was a steaming puddle of melted phone lol Heat would be huge issue, imo.
I used to play some of my apps on it while its charging (I have since stopped lol) which made it fit the catagory of handwarmer for winter. Even smelled hot with an underlying waft of really hot circuitry.

Now, making a clienty app that could give you access to the clients you have running via an account hookup, or something useful but not phonicidal would be nice. Or... a screensaver app for your phone that mimics your data crunching screensaver which would tell you how much time is left on your packets (the information could be shared between your phone app and your client)

I wish I could write programs lol

Nice to see HAL still here :) Doubt he remembers me from way back when tho xD
I first started being crunchy around 1998 or so on and off.
Now I have a machine that can actually do something so I am back as of yesterday.
I has a MiniCity :)
http://en-ki-du.myminicity.com
ID: 1232654 · Report as offensive
-BeNt-
Avatar

Send message
Joined: 17 Oct 99
Posts: 1234
Credit: 10,116,112
RAC: 0
United States
Message 1232996 - Posted: 18 May 2012, 20:26:49 UTC

Until a phone can reach a point of doing something semi-productive, not affect my battery life or stability of my phone it will not be crunching anything. My phone is not a toy to crunch on it's a life line that's only as good as the battery lasts. As bad as some people really want all these massive numbers crunching away at Seti et al, the truth is desktop are so much more powerful and aren't a life line for a lot of people. You don't have to worry about battery power and all the other considerations on a smart phone.

If I had to take a wild guess (I have a Galaxy Nexus) if I could crunch as is right now on it, it would probably pull about a 250rac, if that. The battery would die in about 30-45 minutes doing it, and I wouldn't have a phone, gpu, utility device in my pocket anymore.

Until it gets to a point that every 10 phones would equal half the rac of a decent computer it just doesn't make sense. This conversation may as well be why our home computers still don't and can't do the same type of job a blue gene etc. does. Not to mention the amount of smart phone users vs. Boinc people would probably be less than 1%.

Another point to bring to the table would also need to be brought up about all these numbers being thrown around. Keep in mind there is a certain amount of marketing spin and biased benchmarking being done to show the 'power' of these new mobile devices. Also keep in mind that activation per day don't mean much either considering they don't account for account deactivations or people moving to new devices. Sure having 300,000 activations a day is awesome and a huge number but doesn't mean much if 299,000 of those are people simply moving to a new device. I swap to new devices every 6-12 months....
Traveling through space at ~67,000mph!
ID: 1232996 · Report as offensive
David S
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 18352
Credit: 27,761,924
RAC: 12
United States
Message 1233221 - Posted: 19 May 2012, 4:12:04 UTC - in response to Message 1232996.  

Until a phone can reach a point of doing something semi-productive, not affect my battery life or stability of my phone it will not be crunching anything. My phone is not a toy to crunch on it's a life line that's only as good as the battery lasts. As bad as some people really want all these massive numbers crunching away at Seti et al, the truth is desktop are so much more powerful and aren't a life line for a lot of people. You don't have to worry about battery power and all the other considerations on a smart phone.

If I had to take a wild guess (I have a Galaxy Nexus) if I could crunch as is right now on it, it would probably pull about a 250rac, if that. The battery would die in about 30-45 minutes doing it, and I wouldn't have a phone, gpu, utility device in my pocket anymore.

Until it gets to a point that every 10 phones would equal half the rac of a decent computer it just doesn't make sense. This conversation may as well be why our home computers still don't and can't do the same type of job a blue gene etc. does. Not to mention the amount of smart phone users vs. Boinc people would probably be less than 1%.

Another point to bring to the table would also need to be brought up about all these numbers being thrown around. Keep in mind there is a certain amount of marketing spin and biased benchmarking being done to show the 'power' of these new mobile devices. Also keep in mind that activation per day don't mean much either considering they don't account for account deactivations or people moving to new devices. Sure having 300,000 activations a day is awesome and a huge number but doesn't mean much if 299,000 of those are people simply moving to a new device. I swap to new devices every 6-12 months....

You've got some good points, but I don't think anyone expects to crunch when running on battery. I keep my phone plugged in whenever possible so that when I do have to walk away from the power I have the maximum amount of battery available.

And you know what? When it's plugged into a computer, it still can't keep itself charged if it's running an app that keeps the screen turned on and does a good bit of processing.

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

ID: 1233221 · Report as offensive
Ianab
Volunteer tester

Send message
Joined: 11 Jun 08
Posts: 732
Credit: 20,635,586
RAC: 5
New Zealand
Message 1233239 - Posted: 19 May 2012, 5:17:44 UTC - in response to Message 1232996.  

Until it gets to a point that every 10 phones would equal half the rac of a decent computer it just doesn't make sense.


And reading up on the new Nvidia chips, with over 1500 CUDA cores, it's going to be a while before this happens. OK that sort of GPU wont be mainstream for a while, but in 5 years it probably will. So what if you have a 4 core cellphone, it's still a couple of orders of magnitude less than a desktop...

Ian
ID: 1233239 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51468
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1233245 - Posted: 19 May 2012, 5:22:20 UTC
Last modified: 19 May 2012, 5:22:52 UTC

I thought this thread had died a reasonable death.......
Smartphones shall not ever contribute a significant amount to the project.
Ever.
They were not designed to be able to.
I suppose, you could get on escam and buy a few hundred decommissioned models, plug them all into a few hundred wall worts to keep them powered, and claim that you have achieved handheld crunching....
Go for it.
Silly discussion, in my opinion.

End of line.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1233245 · Report as offensive
-BeNt-
Avatar

Send message
Joined: 17 Oct 99
Posts: 1234
Credit: 10,116,112
RAC: 0
United States
Message 1233268 - Posted: 19 May 2012, 5:52:24 UTC - in response to Message 1233245.  

I thought this thread had died a reasonable death.......
Smartphones shall not ever contribute a significant amount to the project.
Ever.
They were not designed to be able to.
I suppose, you could get on escam and buy a few hundred decommissioned models, plug them all into a few hundred wall worts to keep them powered, and claim that you have achieved handheld crunching....
Go for it.
Silly discussion, in my opinion.

End of line.


Yeah I think the same way for the most part. I could see the combined influence it could achieve but I don't think it would ever be anything more than a "see if we can" project more than anything truly useful.
Traveling through space at ~67,000mph!
ID: 1233268 · Report as offensive
Previous · 1 . . . 6 · 7 · 8 · 9 · 10 · 11 · 12 . . . 13 · Next

Message boards : Number crunching : Smartphone crunching


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.