Posts by Shaktai
|
1)
Message boards :
Number crunching :
Pre-combined Package 64bit SSSE3x
(Message 840155)
Posted 15 Dec 2008 by  Shaktai
Post:
Thanks. I had found the same thing via Google, but appreciate your effort. Was able to extract them successfully. |
|
2)
Message boards :
Number crunching :
Pre-combined Package 64bit SSSE3x
(Message 840144)
Posted 15 Dec 2008 by Shaktai
Post: I have them completed and up on my site as well. Very handy and you have most of the popular platforms covered. Found packages for my Mac Mini core duo, iMac Core2 duo and of course the new Core i7. UPDATE: Hmme, the extractor for Win 64 versions appears to be 32 bit and won't run on my 64bit. Thanks. |
|
3)
Message boards :
Number crunching :
Optimized Astropulse for Windows issues, install, questions
(Message 824720)
Posted 29 Oct 2008 by Shaktai
Post: FOR MAC OS USERS Yes it is. I guess I should have clarified too, that would be about 845 credits per day, per processing core on my iMac C2D 3.06. That compares to around 1300 or more per day, per core for Alex's MB app. (estimated as I have far too many "pending credits" to have reached maximum output yet). Bit improvement but not close to the Multi-Beam app. Thanks to Crunch3r for his work though. It is still a big help. |
|
4)
Message boards :
Number crunching :
Optimized Astropulse for Windows issues, install, questions
(Message 823426)
Posted 26 Oct 2008 by Shaktai
Post: FOR MAC OS USERS After trying Crunch3r's optimized Mac app for Astropulse I saw a significant improvement over the original default app. However it is still well behind AK's Seti enhanced app in credits per day. On my iMac dual core 2 duo 3.06, it took about 21.5 hours to complete and claimed 756 credits. That would be approx 845 credits per day. It started out indicating about 11 hours to complete, but took much longer. |
|
5)
Message boards :
Number crunching :
Optimized Astropulse for Windows issues, install, questions
(Message 822535)
Posted 24 Oct 2008 by Shaktai
Post: Howdy, Thanks. I'll give it a try on my iMac 3.06 Core 2 Duo. Haven't got any work units yet but if I got my app info setup correctly to run alongside MB then all should start getting some units. Had to reinstall BOINC to correct permissions, but that is no big deal. (easier then going into the terminal to correct them) |
|
6)
Message boards :
Number crunching :
Returning to seti, Upgraded to Boinc 6.*, where to place optimized?
(Message 815247)
Posted 6 Oct 2008 by Shaktai
Post: The following thread has all the latest SETI Expanded optimized clients. There is no optimized astropulse client yet. Scroll to the bottom of the thread to get the optimized clients for your platform. (Windows, Linux or Mac OS, etc) http://setiathome.berkeley.edu/forum_thread.php?id=31810 If you are not running Windows or Linux, there is another thread that contains astropulse "ports" for other platforms and OS , but these are ports of the code, not optimized clients. In the thread you can also find instructions on how to modify the app_info.xml file for running both. http://setiathome.berkeley.edu/forum_thread.php?id=48254 Welcome back. I just returned recently myself. |
|
7)
Message boards :
Number crunching :
On which APPLE are you working ?
(Message 710154)
Posted 9 Feb 2008 by Shaktai
Post: Mac Mini Intel 1.66 dual core. |
|
8)
Message boards :
Number crunching :
A HAPPY NEW YEAR to All @ The NC . . .
(Message 696426)
Posted 1 Jan 2008 by Shaktai
Post: May you have a Happy and prosperous New Year!!! Manigong Bagong Taon Feliz Año Nuevo 2008 a great year for everyone and for SETI research. |
|
9)
Message boards :
Number crunching :
Welcome to the 9 Year Club
(Message 592583)
Posted 25 Jun 2007 by Shaktai
Post:
Hey Byron. Wow, that long since I posted??? He He. Guess time does fly. Been around since then but I guess not much to say. Lot's of time on other projects but was around SETI in the early BOINC days. Take care. |
|
10)
Message boards :
Number crunching :
Welcome to the 9 Year Club
(Message 592174)
Posted 24 Jun 2007 by Shaktai
Post: Started with SETI 16 Jun 1999. Haven't been continuous as have worked a lot of other projects as well, but back for now. |
|
11)
Message boards :
Number crunching :
OS X Universal Binary
(Message 230038)
Posted 12 Jan 2006 by Shaktai
Post: Not only will you have to see the BOINC Client as a universal binary you will also have to see the science applications too ... at least for the near future Steve did it to you and I big time ... worse, we will no longer see the advantages of Altivec for Einstein@Home (for example)... Can't blame it all on Steve. IBM didn't exactly do their part in meeting expectations. It was a business decision, and maybe a good one considering the lack of progress from IBM and Moto. Only time will tell. Still, while we loose altivec, the new development tools should help make it possible to take advantage of the Intel chip, and the upsides to Intel's new technology as well, not the least of which is lower power consumption. At least we can hope. Life is always a compromise, but we still have OS-X and it keeps getting better. |
|
12)
Message boards :
Number crunching :
OS X Universal Binary
(Message 229918)
Posted 12 Jan 2006 by Shaktai
Post: Now that the Intel Macs are announced, when will we see BOINC for OS X as a universal binary? New iMacs are availble now (though supply may be limited due to demand). The new MacBook Pro (Powerbook replacement) in February sometime. Both use the new low power dual-core intel chips. Applestore (website) shows shipping in 1-3 days for the iMacs and just "February sometime" for the MacBook Pro. |
|
13)
Message boards :
Number crunching :
Questions about Boinc on new Intel based Macs
(Message 229913)
Posted 12 Jan 2006 by Shaktai
Post: Should be able to compile without much difficulty but might take someone with access to an Intel machine to tweak it right. Not sure what optimizations are available but should be the usual SSE, SSE2 and SSE3. Intel just released some additional tools to developers so someone should be able to do it. Don't know when Berkeley might create an "official" client for it though. In theory, they should be able to compile a "universal" binary that will run on Power PC and Intel using Apple's and Intel's current developer tools so that they only have to maintain one version. Third parties could then tweak the client and the apps for maximum performance. Even without, BOINC should run under Rosetta (the Mac software not the project) but may be a bit slower. Hope so anyway, I want to get one of the new dual core Intel iMacs for my "photo" machine. |
|
14)
Message boards :
Number crunching :
BOINC and Personal Firewalls
(Message 208082)
Posted 9 Dec 2005 by Shaktai
Post: Mac OS-X firewall - 2 computers PC-cillan Internet Security firewall - 2 computers V-Com System Suite 6 firewall - 1 computer. All have the same issue that they can't upload to SETI, except once in a great when one or two units will succeed (maybe 2 or 3 a day will upload). Don't think it is a firewall problem. Most likely the SETI servers. Don't have any problems with any other BOINC projects. These are the messages I get. Three different messages with three different work units all in about the space of three minutes. These three were all from one of the Macs. Fri Dec 9 09:39:21 2005|SETI@home|Temporarily failed upload of 15oc03aa.1719.23856.334664.85_1_0: system I/O Fri Dec 9 09:40:57 2005|SETI@home|Temporarily failed upload of 21mr05ab.11361.20482.698572.175_0_0: timeout Fri Dec 9 09:42:30 2005|SETI@home|Temporarily failed upload of 15oc03aa.1719.18721.392322.46_0_0: error 500 |
|
15)
Questions and Answers :
Macintosh :
PLEASE DO SOMETHING TO MAKE SETI@HOME ENJOYABLE FOR MAC USERS
(Message 191791)
Posted 22 Nov 2005 by Shaktai
Post: Substantial efforts were made both by Berkeley and by other third parties including many Mac users to compile the code for 10.2.8. Unfortunately they all met with little or no success. It would require a modified code base just for 10.2.8, and it is not pratical to maintain a seperate code base just for one OS. BOINC and all associate projects require 10.3.x or greater. It is sad, but true. This has been discussed many times. In the end it isn't a BOINC or SETI argument, but a choice people must make. To stay behind or to move ahead, not just in one area, but all areas. I based my upgrade decisions not on BOINC or SETI, but on every other aspect of Mac use. The benefits were substantial. Being able to run BOINC projects is just an extra frill. |
|
16)
Questions and Answers :
Macintosh :
Installing BOINC on an external drive Maxtor
(Message 164703)
Posted 8 Sep 2005 by Shaktai
Post: The BOINC GUI looks for the file to be in Library on the boot drive. I suspect you could alter some resource files to get it to look elsewhere, but I would have no idea how. However, with the BOINC CLI (if you are comfortable with the terminal) you can run it anywhere. |
|
17)
Questions and Answers :
Macintosh :
? about OSX client version...
(Message 164702)
Posted 8 Sep 2005 by Shaktai
Post: BOINC 4.43 is the BOINC client, not the SETI app. BOINC is just the framework for many different projects and only handles the overhead processes like upload, download and results display. The SETI app itself is the 4.18. That is what does the actual processing. |
|
18)
Message boards :
Number crunching :
Pentium D Processors?
(Message 163117)
Posted 4 Sep 2005 by Shaktai
Post: Maybe someone can enlighten me here, I am planning on building a new cruncher in the future, and being an AMD fan, looking at an athlon 64 X2. I have looked at newegg.com and there seems to be a big dif in prices between the AMD duel cores and the intels. Why is this? Is it supply and demand? The supply chain? What am i missing? There is a difference, but in more then price. Intel has a low cost 820. There really isn't a direct counterpart on the AMD side. The AMD 64 X2 3800 really compares more against the Pentium D 830 for overall general performance. So the price difference isn't that great at that point. My AMD 4200 compares very favorably against my Pentium D 840, and actually has a slight performance edge. At this point, the D840, the cost is now comparable. The AMD 4800 will outperform the D840 EE in most areas. Again, not a big price difference. Intel basically kludged together the Pentium 4 to make their first dual cores quickly, the goal to beat AMD to market. For a "kludge" the Pentium D is a pretty good chip. AMD designed their dual cores from the ground up. (Higher development costs) The end result however was some very efficient chips. A factor to keep in mind with the current chips is also the energy usage. Despite higher initial costs, the ROI (return on investment) will be much better with the AMD's due to significantly lower energy consumption. The AMD's run cooler and use less electricity to provide comparable performance. Of course, these are essentially first generation efforts for both companies. Next year is when the dual core market will become really competitive, as both companies continue to refine their products and bring out new innovations. If you're looking for a dedicated cruncher, it depends on the project as to which chip will do the best job for the dollars. See my prior posts. Different projects do different types of calculations. SETI favors Intel with big L2 caches. Einstein and Predictor favor AMD. (Windows boxes only, I am not considering Macs or other platforms here). The Pentium D is a good chip, but a bit of an energy hog. The AMD is pricier, but more effecient. I own one of each. Were I to add another box, I would go with the AMD at this time. Intel has some interesting stuff in their product pipeline come next year though. If you are looking for general usage, or also a gamer, then the AMD wins hands down on value and performance(with a decent graphics card). AS a dedicated cruncher with the best cross project performance, I would recommend the AMD 4400 or 4800 if you can afford them. The extra cache will help on some projects. The 4200 is a great chip though, with a definite peformance booost over the 3800 and not that much of a higher price. Hope that helped. |
|
19)
Questions and Answers :
Macintosh :
Preferences and Suspension?
(Message 162632)
Posted 3 Sep 2005 by Shaktai
Post: The preferences screen in Mac OS-X is just a default. It is not used by BOINC Manager. Your preferences are actually set through your online account. However, if you go to the Projects tab, Highlight SETI. On the left hand side, you will see a section labeled websites. (you will have to open the window up fairly large) You will then see links for both your account and your SETI preferences. Clicking on those links will take you directly to your account web page where you can change project preferences. I know this seems a little convoluted if you are only running one computer, but for those folks that run multiple computers it is a great, as it allows you to manage preferences from a single interface. You can have up to three different preference settings for both general and SETI (Project specific) preferences, if you use the Home, Work and School settings. The names aren't important, but they allow up to three different settings, and then in the computer list you can designate which settings you want a particular computer to use. NOTE: General settings affect all projects, if you are running other projects along with SETI. Hope that helps. |
|
20)
Questions and Answers :
Macintosh :
Faster core client needed
(Message 162182)
Posted 2 Sep 2005 by Shaktai
Post: Martin, the optimized SETI apps and clients will work with the GUI apps. I run them that way and with multiple projects. To use optimized BOINC Client. Got to apps folder Control click on BOINC icon and select "Show contents". A window will open. double click on contents. Open Resources folder. Place optimized BOINC client in resource folder and change name to "boinc". (you will have remove the old client) Rerun benchmarks. To use the optimized SETI Apps. Go "Hard Drive" > Library > Application Support > BOINC Data > Projects > Setiathome.berkeley.edu. Place the optimized app and app_info.xml file in that folder. Close everything up and restart. |
Next 20
©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.