climateprediction.net home page
Posts by old_user35795

Posts by old_user35795

1) Questions and Answers : Macintosh : Mac support (Message 28984)
Posted 26 May 2007 by old_user35795
Post:
How great to find Mac support again! I came by the website today, just to see what was going on and never suspecting I\'d find a new Mac client. But wow, there it is. I immediately upgraded the old unused BOINC on my drives and my four modest cores are already crunching. Hooray. Thanks guys.
2) Questions and Answers : Macintosh : The Mac Donation - New Mac Support (?) Thread (Message 24372)
Posted 21 Sep 2006 by old_user35795
Post:
As a matter of curiosity, do you have stats for Mac numbers/percentage of total machines in SETI, as well as the top-ranked machines?


Oh, it\'s on boincstats.com. Mac users are a definite small minority. It\'s the 4th most common operating system: XP (760k OS\'s), Win2000 (120k), Linux (82k), Mac (48k), Win2003 (35k), etc...

I\'ve been tracking the performance of my iMac Core 2 Duo (2.16ghz) against other machines. On SETI@Home each work unit gets done 3 or more times... so you can see how long other CPUs take (although I think it\'s not an absolute comparison if the other CPUs are throttled to only give x% to SETI).

Anyway, just anecdotally, I\'ve done 60 work units of various lengths over the last couple of days and on none of them has a Wintel box even come close to the iMac. In fact, the iMac has been the fastest on every work unit so far except one ... and that was a Mac Pro that crunched it faster.

Typical results for the same work units:

iMac: 8,200 seconds CPU time. Linux Dual P4 3.0 ghz box: 29,670 seconds.
iMac: 6,800 secs. XP Athlon 3400+ box: 17,530 secs.
iMac: 7,990 secs. Win 2003 P4 1.6ghz box: 23,750 secs.
iMac: 7,160 secs. XP Athlon Dual-Core 4600+ box: 13,640 secs. XP P4 2.6ghz box: 27,550 secs.

So nobody gets me wrong ... I\'m not picking a PC fight.

My point is simply that the crunching capacity of the Mac base that\'s out there is getting *much* better, hence, they would be a lot more productive for CPDN.

But I just love CPDN, that\'s why I want them to support Macs again. I don\'t want to fold proteins and have my CPU ultimately contribute to the intellectual property of some pharmaceutical company or something like that. CPDN is just far and away the most compelling distributed computing project.

Hey - didn\'t Richard Branson announce jillions for climate change study today? Does he like Macs? :-)
3) Questions and Answers : Macintosh : The Mac Donation - New Mac Support (?) Thread (Message 24363)
Posted 20 Sep 2006 by old_user35795
Post:
If one looks at the top computers at SETI@Home now, six of the top ten are Macs. And those are mainly G5\'s. As the MacPros and Core 2 Duo iMacs get up to speed, they will be rising through the ranks. I\'m sure there will be latest and greatest PCs coming too; but I think it\'s a virtual certainty now that an attempt to accomodate Mac users would be more \"profitable\" for a distributed computing project than in the past.

Has anyone pinged the people at CPDN about Mac support lately ... or should the absence of a reply here be taken as a lack of interest?
4) Questions and Answers : Macintosh : The Mac Donation - New Mac Support (?) Thread (Message 23910)
Posted 11 Aug 2006 by old_user35795
Post:
Oh man... is it that bad? No interest at all?
5) Questions and Answers : Macintosh : The Mac Donation - New Mac Support (?) Thread (Message 23877)
Posted 8 Aug 2006 by old_user35795
Post:
All,

Okay, as a last-ditch effort, if there are still enough people checking this board, can we discuss the possibility of getting donated Mac hardware to CPDN?

I find climateprediction.net to be far and away the most interesting distributed computing project. It\'s far more important and compelling than, say, ... okay, I won\'t insult anybody else .... those other projects.

I shed tears when CPDN suddenly pulled the plug on Mac support. Combing through the posts here, it seems like the biggest impediment to renewed Mac support is the lack of Mac hardware for the developers to use in Oxford. So, my suggestion is that we consider pooling our resources to solve that problem. Let\'s give them the Mac(s) that the developers want.

Of course there are a lot of questions, among them:

1) Is CPDN interested in this proposition?

2) Are other (former) Mac CPDN users interested?

3) If we got them Mac(s), how firm a commitment can CPDN make that Mac crunching will once again be supported?

4) What kind(s) of Mac(s) are necessary? (And what models could be supported ... I think I am hearing Intel on this one...)

5) If we move forward, should we focus on collecting contributions for buying CPUs, or on lobbying Apple (or somebody else) to give them what they need, or on some combination of the two?

I guess questions 4 and 5 are getting a little ahead of the game. Could somebody from CPDN weigh in here with some thoughts? Or could a moderator contact them to ask? If we got them the hardware, will they have the time and ability?

A show of hands from other users? Is there interest? Are there enough of us left? Can we do this?


6) Questions and Answers : Macintosh : Inducement to CPDN Programmers: Mac Pro (Message 23876)
Posted 8 Aug 2006 by old_user35795
Post:
Donate one to Oxford Uni, and you MIGHT get an app.


Ha. Well, despite all the unseemly CPDN grousing about how small and unfunded they are, they are much bigger and probably better paid than my little nonprofit. So, not much chance of a donation of that size from here.

The last new Mac I bought was a Blue and White G3. That was about 1999. This Mac Pro will probably still be in service in 2015, held together by paper clips and used chewing gum. Such is life at The Sunshine Project.

Anyway, as far as donations go, check out the Mac Donation thread I started. We should discuss this. I hope others agree.



7) Questions and Answers : Macintosh : Can\'t get work for my Mac (Message 23872)
Posted 8 Aug 2006 by old_user35795
Post:
And, at present, the programmers don\'t have a Mac on which to build and test.

Which Mac would you like?


Good question.

A modest proposal: If the CPDN people could assure us that they would work to produce a current Mac client and could commit to maintaining it for a reasonable period, then couldn\'t the Mac CPDN people band together and get them an Intel Mac?

What are the system requirements? The Mac Pro\'s are pretty pricey ... would an Intel iMac or even a Mac Mini do? Is any Intel machine enough ... or would you need both a Core Duo and Xeon?

An additional thought for people in the US. Given the nature of CPDN, it would probably not be too hard to locate a nonprofit organization that could offer tax deductibility for any donations ... provided this hypothetical nonprofit and CPDN could sign a simple understanding in which CPDN assured that the hardware would be used for its nonprofit research purpose.
8) Questions and Answers : Macintosh : Inducement to CPDN Programmers: Mac Pro (Message 23871)
Posted 8 Aug 2006 by old_user35795
Post:
Hey - I\'m just about ready to order a Xeon 2.66ghz quad processor Mac. Wow, think of all the CPU cycles those otherwise idling extra processors could dedicate to CPDN ... :-)

Anybody else want to dedicate one of these monsters to CPDN? ... Ah, if only there was a Mac client...
9) Questions and Answers : Macintosh : Feeling betrayed.... (Message 22077)
Posted 16 Apr 2006 by old_user35795
Post:
I know the CPDN programmers have a lot to worry about; but, yikes, I spent a lot of CPU time on this. It\'s the only distributed computing project worth a d*mn. What a shame. I would have at least expected an e-mail or some sort announcement / thank you from the CPDN team before they pulled the Mac plug.
10) Questions and Answers : Macintosh : What about this crash (segmentation fault)? (Message 17058)
Posted 8 Nov 2005 by old_user35795
Post:
Is there anyting that can be done about this? Or do I toss the model and start again? (I tried restarting it ... it rewound 30 steps and then crashed at the same point again.)

sulphur_47ye_000297206 - PH 1 TS 0162030 A - 16/04/1820 15:00 - H:M:S=0591:28:42 AVG=13.14 DLT= 6.90
SIGSEGV: segmentation violation
Crashed executable name: boinc
built using BOINC library version 5.2.5
System version: Macintosh OS 10.4.3 build 8F46
Stack Frame backtrace:
# Flags Frame Addr Caller PC Symbol
=== === ========== ========== ==========
1 --- 0xbfffdd30 0x90011b28 ___vfprintf + 0x14e8
2 --- 0xbfffe3e0 0x9001061c _sprintf + 0xfc
3 --- 0xbfffe580 0x0002a228
4 --- 0xbfffe7e0 0x0002a234
5 --- 0xbfffea40 0x0002a234
6 --- 0xbfffeca0 0x0002a234
7 --- 0xbfffef00 0x0002a234
8 --- 0xbffff160 0x0002a234
9 --- 0xbffff3c0 0x0001a610
10 --- 0xbffff8c0 0x00023f20
11 --- 0xbffffa20 0x0001af90
12 --- 0xbffffa70 0x0000d404
13 --- 0xbffffae0 0x00022898
14 --- 0xbffffb30 0x0002294c
15 --- 0xbffffb80 0x00006cc0
16 --- 0xbffffbe0 0x00006b34
17 -P- 0xbffffc20 0x00000000
18 FP- 0x00000000 0xffffffff

Exiting...
logout
[Process completed]
11) Questions and Answers : Macintosh : Upgrading the BOINC client... worth it? easiest route? (Message 10786)
Posted 13 Mar 2005 by old_user35795
Post:
The climateprediction.net server has been enticing me to upgrade to version 4.19 of the BOINC client for several days (when it trickles, I get that message). I've ignored it until now, because I was well into a work unit and didn't want to upset anything.

Is switching to the new version worth it for me or for climateprediction.net? Or should I hang tight with my happily functioning 4.13?

If I should upgrade, instructions on how to do so (without killing the WU in progress) for the command-line-challenged would be appreciated.

12) Questions and Answers : Macintosh : Power Mac G3 and being given workunits (Message 10785)
Posted 13 Mar 2005 by old_user35795
Post:
I wouldn't if it was my primary machine; but if it's parked in a corner and crunching, let it run. Why not? (Says the guy with a quasi-abandoned Power Computing 604/150 in the corner doing SETI at home.)

Evan, I say go for it.
13) Questions and Answers : Macintosh : System crashed, on restart BOINC downloads new model, won't work on old one (Message 7320)
Posted 16 Jan 2005 by old_user35795
Post:
I downloaded my first model and was happily crunching for a few days, up to about timestep 21,000. My system (G4 tower, 10.3.7) froze for reasons unrelated to BOINC.

After rebooting, I restarted BOINC; but it downloaded a new model rather than resuming the old one. It also interspersed the new model's files in with the old, creating a new projects directory inside the 'jobs' directory of the old one, among other things.

I eventually tossed the entire project directory and restarted BOINC. It downloaded a third model and everything is fine again, although I guess the 21,000 timesteps already done were pointless.

For future reference, can somebody please let me know how to get BOINC to resume the old model after a crash? It would be appreciated if this were explained in terms comprehensible to somebody who doesn't typically use a command line interface.

EH




©2024 climateprediction.net