Posts by Keck_Komputers
|
21)
Message boards :
Number crunching :
AMD Phenom
(Message 743880)
Posted 25 Apr 2008 by  Keck_Komputers
Post: Is it possible for two motherboards with the same model name (M2N68-LA), to have different sockets? HP seems to think so.... I would hate to have my own little computer shop, and order a dozen motherboards, all the same, only to find out when they got here, that half of them dont have the same socket.... Thats messed up. But according to what HP is telling me, that is the way it is. (Why am I having a hard time believing this?)It might be that there is a BIOS update required to recognize more than 2 cores. In that case their response would make some sense, although telling you to download a new BOIS would make more sense than just telling you it would not work. |
|
22)
Message boards :
Technical News :
Walkthrough (Apr 22 2008)
(Message 743451)
Posted 24 Apr 2008 by Keck_Komputers
Post:
Correct, there is often more data collected in a day than will fit on a "tape". With those names there should be at least eight "tapes" used on that day, the first one for the day is normally "aa". |
|
23)
Message boards :
SETI@home Science :
Seti Classic Poll
(Message 739700)
Posted 16 Apr 2008 by Keck_Komputers
Post: You can get close to the look how to |
|
24)
Message boards :
Number crunching :
To Screen-save or not to screen-save?
(Message 739694)
Posted 16 Apr 2008 by Keck_Komputers
Post: Is it even possible to "save" LCD displays? Yes. Screen savers are not as needed as they were years ago. Plasma screens need savers just like the old days. Other display types can still suffer from burn in however it takes so long now that it is not worth mentioning. |
|
25)
Message boards :
Number crunching :
To Screen-save or not to screen-save?
(Message 739280)
Posted 15 Apr 2008 by Keck_Komputers
Post: Studies made in beta showed that running the screensaver increased run times less than 1% on systems with separate openGL compatible cards. There is a preference in the SETI specific settings that allow you to limit the amount of CPU time that will be used to process graphics. |
|
26)
Message boards :
Number crunching :
Wait a second... How is Allen Telescope Array data crunched??...
(Message 738791)
Posted 14 Apr 2008 by Keck_Komputers
Post: Are you sure about that? SETI@home is just one small part of the SETI efforts. Most SETI surveys look for relatively strong signals, this is the type of survey they are expecting to use the ATA for. I do not know if the data collected there will be compatible with SETI@home. Since Arecibo is a single dish it can receive fainter signals. To illustrate; if you have a bunch of sails spread over a large area you have a better chance of catching some wind however a single sail will be able to convert more wind energy into motion in the right conditions. Since the ATA is composed of a bunch of small dishes it could theoretically point each one differently and have as many "beams" as there are dishes. However each "split" reduces sensitivity so three is probably the limit where loss of sensitivity balances the increase in "beam width". |
|
27)
Message boards :
Number crunching :
boinc resource share
(Message 732990)
Posted 31 Mar 2008 by Keck_Komputers
Post: If your client does not have at least 10 days of work on hand (your CI) it will download work from wherever it can get it. If it has less than 20 days of work on hand (CI + AQ) this behaviour may persist. Because E@H has 14 day deadlines your queue will never meet both of these conditions and rarely meet the 10 day CI. So any time the client connects to report E@H work it will top of the queue while it has the database open. Any time the CI is more than 40% of the shortest deadline of any project attached you will have problems like this one. Since you are attached to SETI the longest recommended CI is 1.8 days because SETI has deadlines as short as 4 days. In addition CI + AQ should not be more than 80% of shortest the deadline or 3.6 days. The one saving grace is that SETI has varying deadlines so you can base your calculations on E@H's consistent 14 day deadline, or 5.6 for the CI and 11.2 for the sum of the two. |
|
28)
Message boards :
Number crunching :
Orphaned Work Units
(Message 731072)
Posted 27 Mar 2008 by Keck_Komputers
Post: Another mystery for you Computer Whizzes. The core and core 2 architectures are much more efficient than the P4. So you get more processing done per clock tick. I am not sure of the exact ratio but it is greater than 2 to 1. So your laptop should be comparable or a bit faster than your desktop per core. |
|
29)
Message boards :
SETI@home Science :
Gravitational Lensing
(Message 730910)
Posted 27 Mar 2008 by Keck_Komputers
Post: That is what this project is all about. Trying to find those weak signals that have been distorted by their journey here. |
|
30)
Message boards :
Number crunching :
Odd graph picture
(Message 728576)
Posted 21 Mar 2008 by Keck_Komputers
Post: There is most likely a bad date in one of your statistics files. Go to your BOINC directory and delete all statistics*.xml files. In a few days you should start getting more realistic graphs. |
|
31)
Message boards :
Number crunching :
Boinc too complex??
(Message 728086)
Posted 20 Mar 2008 by Keck_Komputers
Post: One idea that I like that I don't see mentioned in this thread, is set the quota to 1 for new hosts. That way the host can download enough work to get started, but cannot download more until some is returned. The max quota would be the same and would be quickly reached if the host is working properly. However there would still be problems with new participants worried about their queue filling. Yep, but as soon as it is downloaded the client will make a full request. |
|
32)
Message boards :
Number crunching :
Boinc too complex??
(Message 727809)
Posted 19 Mar 2008 by Keck_Komputers
Post: One idea that I like that I don't see mentioned in this thread, is set the quota to 1 for new hosts. That way the host can download enough work to get started, but cannot download more until some is returned. The max quota would be the same and would be quickly reached if the host is working properly. However there would still be problems with new participants worried about their queue filling. |
|
33)
Message boards :
SETI@home Science :
Meteor falls near Spokane
(Message 720625)
Posted 2 Mar 2008 by Keck_Komputers
Post: I think you are right on for one of the prime reasons of horizon flashes. Another is that the object is not going to flash/break up immediately on contact with the atmosphere, it will take some time before the stresses build up enough to cause it. So the object will be closer to the surface and therefore closer to the horizon for more observers. |
|
34)
Message boards :
Number crunching :
UPdate to 5.10.43 , better graph display, also better performance?
(Message 720213)
Posted 1 Mar 2008 by Keck_Komputers
Post: Sometimes the graphics line up on their own. They will get off center again eventually. I have seen the shifted graphics in version 6.00 already. |
|
35)
Message boards :
SETI@home Science :
Meteor falls near Spokane
(Message 719414)
Posted 28 Feb 2008 by Keck_Komputers
Post: Youtube has a ton of really cool meteor(ite) videos. One phenomena that I've noted is that almost all the meteor(ites) seem to "flare up" intensely right before passing below the horizon. The "flare ups" don't appear to be from impacts. I'm not even going to speculate on this phenomena, except to say that it is interesting. As the meteor heats up it is likely to shatter. If/when that happens there is more surface area to "burn" so you see a "flare up". |
|
36)
Message boards :
Number crunching :
PIV Hyperthreading question?
(Message 718985)
Posted 27 Feb 2008 by Keck_Komputers
Post: It will be interesting to see how well Intel addressed the shortcomings with HT when they rollout the 2nd generation of it in Nehalem. I wonder if they even recognise any shortcommings with HT. It was designed to find hidden performance in the CPUs and it did find about 15% to 30%. |
|
37)
Message boards :
Number crunching :
optimized BOINC
(Message 716308)
Posted 21 Feb 2008 by Keck_Komputers
Post: There is very little advantage to an optimized client, an optimized science app can be a big help though. I don't recommend them since they must be updated manually however I am sure someone will be along with a link soon. |
|
38)
Message boards :
Number crunching :
CPU affinity in multicore systems
(Message 716307)
Posted 21 Feb 2008 by Keck_Komputers
Post: An interesting debate, and concensus is there is no advantage to using processor affinity. I would say the advantage is negligable rather than none existant. Earlier comparisons seemed to indicate there was some advantage to setting affinity with discreate processors, less than 5% though. Since there are so few systems of this type it was not the effort to put into a standard client. |
|
39)
Message boards :
Number crunching :
Redundant result?
(Message 716301)
Posted 21 Feb 2008 by Keck_Komputers
Post:
Yeah we might need to look into the secure access code some more. I have gotten a few WCG tasks since updating to the 6.1.x clients, however none of them have successfully returned and validated. Even though occasionaly things get through enough is blocked to prevent it. |
|
40)
Message boards :
SETI@home Science :
Quick question about the multi-beam receiever
(Message 715565)
Posted 20 Feb 2008 by Keck_Komputers
Post: That's right. The data is also recorded at a higher resolution, 4 bit instead of 2 bit. |
Previous 20 · 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.