Message boards :
Number crunching :
Seti BOINC SAH WU Times
Message board moderation
Author | Message |
---|---|
Sir Ulli Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 |
my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) Athlon64 3.200+ ca 2:30 for one WU Athlon 2.500+ ca 3:10 for every WU so no can beat the P4s, also no Pentium M Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © |
Chilean Send message Joined: 6 Apr 03 Posts: 498 Credit: 3,200,504 RAC: 0 |
|
Neil Walker Send message Joined: 23 May 99 Posts: 288 Credit: 18,101,056 RAC: 0 |
<blockquote>my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) Athlon64 3.200+ ca 2:30 for one WU Athlon 2.500+ ca 3:10 for every WU so no can beat the P4s, also no Pentium M </blockquote> Errm ..... my Athlon64 seems to be beating that at around 72 mins on average. :P Be lucky Neil |
maoshouse Send message Joined: 14 Aug 04 Posts: 12 Credit: 50,639 RAC: 0 |
<blockquote>my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) Athlon64 3.200+ ca 2:30 for one WU Athlon 2.500+ ca 3:10 for every WU so no can beat the P4s, also no Pentium M Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © </blockquote> you must not be doing something right on your computers. my AthlonXP-M is 1wu per 2:08. |
Ingleside Send message Joined: 4 Feb 03 Posts: 1546 Credit: 15,832,022 RAC: 13 |
<blockquote>my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) Athlon64 3.200+ ca 2:30 for one WU Athlon 2.500+ ca 3:10 for every WU so no can beat the P4s, also no Pentium M </blockquote> Well... your p4 uses over 2 hour on one wu, so isn't the fastest. ;) If you starts calculating per cpu (real or ht), the fastest computer in SETI@home/BOINC uses 6 minutes/wu, roughly 240 wu/day. In "classic", the fastest platform was... alpha-compaq-T64Uv4.0d/EV67 there average cpu-time is 1 hr 00 min 00.4 sec As for fastest cpu this is alpha with average cpu-time 0 hr 59 min 25.2 sec |
Heffed Send message Joined: 19 Mar 02 Posts: 1856 Credit: 40,736 RAC: 0 |
<blockquote>my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) </blockquote> HeHe... Not exactly... It means every 2:35 a new WU. :) |
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0 |
<blockquote>If you starts calculating per cpu (real or ht), the fastest computer in SETI@home/BOINC uses 6 minutes/wu, roughly 240 wu/day. </blockquote> Where did you see this? |
Sir Ulli Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 |
<blockquote><blockquote>my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) Athlon64 3.200+ ca 2:30 for one WU Athlon 2.500+ ca 3:10 for every WU so no can beat the P4s, also no Pentium M </blockquote> Errm ..... my Athlon64 seems to be beating that at around 72 mins on average. :P </blockquote> you are using an optimized Client for Linux is that right is Linux 2.6.11-gentoo-r7 an 32 or an 64 Bit OS Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © |
Neil Walker Send message Joined: 23 May 99 Posts: 288 Credit: 18,101,056 RAC: 0 |
<blockquote> you are using an optimized Client for Linux is that right </blockquote> Yes. <blockquote> is Linux 2.6.11-gentoo-r7 an 32 or an 64 Bit OS </blockquote> It's either, depending on how you compile it. In this particular case, it's 64 bit. Be lucky Neil |
Ingleside Send message Joined: 4 Feb 03 Posts: 1546 Credit: 15,832,022 RAC: 13 |
<BLOCKQUOTE><blockquote>If you starts calculating per cpu (real or ht), the fastest computer in SETI@home/BOINC uses 6 minutes/wu, roughly 240 wu/day. </blockquote> Where did you see this?</BLOCKQUOTE> A look on top computers-listing by RAC, shows a 64-cpu-box... The cpu-time seems to be a little more than 6h/wu, with 64 cpu's this gives 6 minutes/wu. 6 minutes/wu equals 240 wu/day. Since the daily quota in seti is 100 per cpu, with upper limit 4 cpu's per computer, this means 400 wu/day as upper limit. Didn't check if really crunches 240 results/day, but this computer have atleast the cpu-power to do it... |
Sir Ulli Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 |
btw have you noticed this my Athlon64 3.200+ does a WU in 2:30 under Windows XP SP1 but when i look here http://setiweb.ssl.berkeley.edu/results.php?hostid=841391 Athlon 3.500+ Wu Times 55xx that means a wu in 1:30 and http://setiweb.ssl.berkeley.edu/results.php?hostid=819359&offset=60 Athlon 4.000+ WU Times around 4xxx that means a Wu in 72 Minutes, that is very fast, so someone is using Linux on Fedora Core 3x86-64 or somethink else..with an optimised Client...you can speak for this and for more info, look at Neds Compiling the seti client under linux x86_64 fantastic Side... Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © |
ampoliros Send message Joined: 24 Sep 99 Posts: 152 Credit: 3,542,579 RAC: 5 |
I'm now a big advocate of compiling clients for Linux, and here's why: I'm running a 2.0GHz P4 (Willamette) with a 512Kb Cache and 400MHz FSB for my "everyday use computer." When I ran SetiBOINC under the Windows partition with the stock clients, I would get times that were OK. Recently I decided to compile my own client and run SetiBOINC under the Linux partitions. I had compiled my own programs before, but it was usually because there were no binaries available for the program that I needed. But what a difference. Now I crunch a WU in 135 minutes. I know that's not a groundbreaking time by any means. I just wanted to say that Metod and Ned inspired me to not settle for stock and to go out and compile my own. 7,049 S@H Classic Credits |
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0 |
<blockquote> A look on top computers-listing by RAC, shows a 64-cpu-box... The cpu-time seems to be a little more than 6h/wu, with 64 cpu's this gives 6 minutes/wu. 6 minutes/wu equals 240 wu/day. </blockquote> Actually, it is still 6 hours per Work Unit, with 64 Work Units done each 6 hours, giving 256 per day. |
Ingleside Send message Joined: 4 Feb 03 Posts: 1546 Credit: 15,832,022 RAC: 13 |
<blockquote> Actually, it is still 6 hours per Work Unit, with 64 Work Units done each 6 hours, giving 256 per day.</blockquote> The cpu-time is 6 hours, but the run-time between each time you reports a result is 6 minutes. |
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0 |
<blockquote><blockquote> Actually, it is still 6 hours per Work Unit, with 64 Work Units done each 6 hours, giving 256 per day.</blockquote> The cpu-time is 6 hours, but the run-time between each time you reports a result is 6 minutes. </blockquote> You are assuming an even distribution of work completion. This is one of the reasons people get confused with BOINC. Each processor within the machine takes 6 hours to complete a work unit, with 4 processors it completes work as I stated above, for, as you stated, a throughput that delivers, on average, one result every 6 minutes. But, the machine is not, in fact, returning a result each 6 minutes. |
Arm Send message Joined: 12 Sep 03 Posts: 308 Credit: 15,584,777 RAC: 0 |
<blockquote>my P4 is the fastest P4 3.2 with HT 2:35 for 2 WUs that means every 77 Minutes a new WU :) Athlon64 3.200+ ca 2:30 for one WU Athlon 2.500+ ca 3:10 for every WU so no can beat the P4s, also no Pentium M Greetings from Germany NRW </blockquote> Err... I saw on your stats the P4 reports an average WU time about 12 000 seconds. That means about 3h 20 mins per WU which is normal for P4 3.2 :) not 77 mins per WU :) |
Ingleside Send message Joined: 4 Feb 03 Posts: 1546 Credit: 15,832,022 RAC: 13 |
<blockquote>You are assuming an even distribution of work completion. This is one of the reasons people get confused with BOINC. Each processor within the machine takes 6 hours to complete a work unit, with 4 processors it completes work as I stated above, for, as you stated, a throughput that delivers, on average, one result every 6 minutes. But, the machine is not, in fact, returning a result each 6 minutes.</blockquote> Oh, did I forget to add average time between a result is finished... BTW, with 64 cpu's my guess is this is already fairly averaged-out if the server runs SETI@home/BOINC 24/7... Atleast from my 1st post it should be apparent it was divided by #cpu's and is therefore an average, if this doesn't come across my exquse is English isn't my strong point so got lost in translation... |
simplify / aki Send message Joined: 30 May 99 Posts: 61 Credit: 2,801,095 RAC: 0 |
<blockquote> but when i look here http://setiweb.ssl.berkeley.edu/results.php?hostid=841391 Athlon 3.500+ Wu Times 55xx that means a wu in 1:30 </blockquote> Feel free to link to my computer :p ;) |
Arm Send message Joined: 12 Sep 03 Posts: 308 Credit: 15,584,777 RAC: 0 |
Impressive! |
longi Send message Joined: 1 Dec 02 Posts: 6 Credit: 1,224,234 RAC: 0 |
Hi P4 look her: http://setiweb.ssl.berkeley.edu/sah/results.php?hostid=712366 Do not Overclock this time. 2 WU in sometimes shorter than 2h. Cu Sometime I use http://boinc.us.tt/ Sometime I use http://mkx.feridot.com/boinc/ Sometime I use http://www.pperry.f2s.com/boinc.htm |
©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.