Kim Vater 的帖子

101) 留言板 : Number crunching : Which is faster? (消息 627957)
发表于:27 Aug 2007 作者: Kim Vater
Post:
What's faster a 3 Ghz intel or a 4000 amd?

This was asked at our forum,and I had no idea so I posted it here.


Google.com is always your friend ;)

Running a "CPU compare" search reveals that 'TomsHardware' has an eminent comparison CPU chart between AMD & Intel processors.

It's not possible to compare new CPU's directly on Ghz anymore because of different CPU architectures.
You need benchmarks like the above mentioned.

Kiva
102) 留言板 : Number crunching : CPU Request (消息 627899)
发表于:27 Aug 2007 作者: Kim Vater
Post:

The case is a 540 watt psu Shinobi SH09A Midi tower. PSU fan and Side fan. Fans are also on the CPU and there is a second fan on the MB. Theory says there should not be an overheating issue.

Total 4 fans.

Greg.


It sounds to me that you don't have sufficient airflow in the midi-cabinet.

You should consider moving the side-fan to front low end of the cabinet or buy another fan to put there.
The trick is to have air flowing past the CPU - not onto it.
A sidedoorfan usualy just make to much turbulence in small midi-cabbis (recycling the hot air).

Edit\\ If it runs cooler with the sidedoor off/open - then you should consider the above mentioned.

Kiva
103) 留言板 : Number crunching : Server stats a-changing... (消息 627460)
发表于:26 Aug 2007 作者: Kim Vater
Post:
The kitties have been looking at the Scarecrow Graphs and they see that the splitters appear to have stopped dead in their tracks, and the purging appears to have slowed or stopped. The assimilator appears to have been out of the action for awhile now, the files must getting rather large.
Is somebody in the shop making adjustments, or are the servers getting tangled again?


Yes - It looks like a busy monday morning for the staff.
More trouble ahead - just as my RAC has leveled out from a steep dive and startet to climb a little. Sad song!

Kiva
104) 留言板 : Number crunching : What's in YOUR rac? (消息 625834)
发表于:24 Aug 2007 作者: Kim Vater
Post:
FWIW, here are a couple of Core 2 Quads running stock 5.27, I presume they were on stock 5.15 prior to the change:

Intel(R) Core(TM)2 Quad CPU @ 2.66GHz RAC 1,219.29
SAH host 3617805 BoincStats.

Intel(R) Core(TM)2 Quad CPU @ 2.40GHz RAC 1,213.75
SAH host 3601179 BoincStats.

The 30 day BoincStats bar graphs for credit per day have too much variation to really judge what effect Multibeam has had, IMO.
                                                                Joe


Those host were added in late July - and cannot be used as an examble! IMHO
According to the very handy site www.andyk.de/boinc/pc.php - these 2 Quads are doing around 100 credits/Hour = 2400 RAC each.

Plz - do not use these 2 rigs for calculating or adjusting the MB multiplier.
Listen to some of the heavy cruncher instead - thanks.

Kiva
105) 留言板 : Number crunching : A little ray of kitty sunshine...... (消息 624092)
发表于:21 Aug 2007 作者: Kim Vater
Post:
He he! :-)

On Saturday I replaced an E6600@3.15Ghz with an Quad6600@3.15Ghz (Stepping G0).
Nothing happend to that steep downhill RAC curve - just increased the pending credits :-(

I'm also at 25-30% decrease in RAC on all rigs - at the moment! (except for a newly added Laptop)

Kiva
106) 留言板 : Technical News : Splitsville (Aug 16 2007) (消息 623568)
发表于:20 Aug 2007 作者: Kim Vater
Post:

There's plenty of crunchers that have been here since the beginning that can tell you about some of the nightmares when SETI Classic had problems. They were just less noticeable then. ;)


I don't think they were less noticeable.
But have in mind that those were the "dial-up modem" days, and I think that all active chrunchers were using thirdparty buffer/cache programs. These progs were very easy to find on the Seti main page at that time.
I used "SetiBuffer" and never ran out of work - but sure had problems connecting to Seti once in a while ;-)

Kiva


107) 留言板 : Number crunching : v5.27 (MB) live on main. (消息 618793)
发表于:13 Aug 2007 作者: Kim Vater
Post:
Hi, I just came back from vacation and had to read a bunch of new threads.

As usual I find Crunch3r's "down to earth/easy to understand" answers very usefull to my worries.

Thanks a lot Crunch3r ;)

(I am now upgrading all rigs to "Chicken Soup 2.4 MultiBeam")

Kiva
108) 留言板 : Number crunching : http error in uploads (消息 609872)
发表于:27 Jul 2007 作者: Kim Vater
Post:
Same problem here in Norway :-(
Started around 16:30 UTC.

Kiva
109) 留言板 : Number crunching : I am back (who? II) (消息 598643)
发表于:4 Jul 2007 作者: Kim Vater
Post:
Who?'s computer are now hidden?? ;)

Kiva
110) 留言板 : Number crunching : I am back (消息 596161)
发表于:30 Jun 2007 作者: Kim Vater
Post:

Anyway, my lowly single-core 3700+ that I have ramped up to FX-57 level speed can easily beat a 3.6GHz Prescott on a per-result basis. Hyperthreading may give the Prescott the RAC advantage over me. Not sure, as I haven't let this system go to its' full potential due to switching to/from Einstein... I think this system would end up with an RAC of 900-1000, and the highest 3.6 Prescott I found is currently at 1,084 RAC...so perhaps I could match RAC with a 3.4... I'll try dedicating this system and see where it ends up...
Brian


@ Brian Silvers
Maybe a bit high exspectations?
I have 2 Prescotts (3.0 and 3.2Ghz) each doing around 1000 RAC.
The 3.2Ghz is doing other task than Seti, and could probably end up with a RAC over 1100 ;-)

Kiva
111) 留言板 : Number crunching : Switching team - what happens with credit? (消息 585388)
发表于:11 Jun 2007 作者: Kim Vater
Post:
@msattler

Thanx for link - you're the man ;) (been searching for a day)

Yes - somethings wrong @ SETI.
Noticed that they booted the servers yesterday (sunday) - but it didn't seem to help much on xternal stat-updates or reporting/downloading WU's.
They'll probably look into it whitin the hour (monday morning CA time)

Happy cruncing

Kiva
112) 留言板 : Number crunching : Switching team - what happens with credit? (消息 585370)
发表于:11 Jun 2007 作者: Kim Vater
Post:
Thanks for the answers so far - but just left me confused now!

Plz don't misunderstand me - it's OK by me, that people cannot take their credits with them to other teams.

I need some kind of more official documentation on the subject, as I plan to write a "How to" guides in the team forums.
I cannot use any "guesing/thinking/might-be-so" posts as documentation in any forums ;-) No offense.

I was hoping somebody could post a link(s) to some public documentation from more "reliable" sources ex. SETI/BOINC staff ?

BTW: I wasn't shifting team in anger or thrown out. Just shifted to a team (within the country) that has an active cruncher-forum and a lot more dedicated DC crunchers (competitors) ;-)

Kiva
113) 留言板 : Number crunching : Switching team - what happens with credit? (消息 585335)
发表于:11 Jun 2007 作者: Kim Vater
Post:
I just recently (4 days ago) switched team from "hardware.no" to "Team Norway".

I show correctly up in different stat-pages as shifted to the new team - but it doesn't seem that my credits (2.2 mill) is following me and is added to the new team (and detracted from old team) ?

Is there some kind of delay with the credits when swithing team? (would be a good thing thou) ?

Kiva
114) 留言板 : Number crunching : S@H just aborted 337 tasks on my machine (消息 583824)
发表于:8 Jun 2007 作者: Kim Vater
Post:
Well I got the same mysterious thing here on 4 out of 7 rigs:

08.06.2007 19:39:29|SETI@home|Sending scheduler request: Requested by user
08.06.2007 19:39:29|SETI@home|Reporting 17 tasks
08.06.2007 19:39:49|SETI@home|Scheduler RPC succeeded [server version 509]
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 09fe05aa.7918.1889.36084.3.223_1 (expired)
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3024.509654.3.242_0
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3024.509654.3.253_0
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3024.509654.3.245_0
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 15mr05aa.5228.16081.848592.3.14_4
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 15mr05aa.455.25681.136088.3.32_4
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 14fe05aa.3729.32689.748586.3.200_3 (expired)
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 15mr05aa.455.26241.429840.3.110_4 (expired)
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 15mr05aa.24057.13393.686074.3.29_3
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 15mr05aa.17676.19488.209664.3.71_3
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3218.204836.3.0_2
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 29ja05ab.2468.3218.204836.3.1_0 (expired)
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3218.204836.3.7_1
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3218.204836.3.3_2
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 29ja05ab.2468.3218.204836.3.14_2 (expired)
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3218.204836.3.13_1
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3218.204836.3.16_0
08.06.2007 19:39:49|SETI@home|Message from server: Resent lost result 29ja05ab.2468.3218.204836.3.22_1
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 29ja05ab.2468.3218.204836.3.23_0 (expired)
08.06.2007 19:39:49|SETI@home|Message from server: Didn't resend lost result 29ja05ab.2468.3793.411056.3.130_0 (expired)
08.06.2007 19:39:49|SETI@home|Deferring communication for 11 sec
08.06.2007 19:39:49|SETI@home|Reason: requested by project

What happens - somebody plz give an plausible explanation ??
None of the apperently expired WU's were near deadlines!!

(The above cut&paste from Boincmng is from my C2D E6600 rig)
Kiva
115) 留言板 : Number crunching : New error cropping up... (消息 577489)
发表于:28 May 2007 作者: Kim Vater
Post:
Had same problem and it's fixed now!

Go to "your account" and push the update-button in your "Seti@home preferences"
Then push the "Update-button" in BOINC manager.

Should fix the problem - at least it did for me ;)

Kiva
116) 留言板 : Number crunching : Uh-oh, what's this about?? (消息 577485)
发表于:28 May 2007 作者: Kim Vater
Post:
Had same problem and it's fixed now!

Go to "your account" and push the update-button in your "Seti@home preferences"

Then push the "Update-button" in BOINC manager.
Should fix the problem - at least it did for me ;)

Kiva



117) 留言板 : Technical News : Chaos at the Greasy Spoon (May 24 2007) (消息 577012)
发表于:27 May 2007 作者: Kim Vater
Post:
Somekind of network/server problem shows at the Gigabit graph ??
http://fragment1.berkeley.edu/newcricket/grapher.cgi?target=%2Frouter-interfaces%2Finr-250%2Fgigabitethernet2_3;view=Octets;ranges=d

Kiva
118) 留言板 : Number crunching : Welcome to the 9 Year Club (消息 576919)
发表于:27 May 2007 作者: Kim Vater
Post:
Today's the day!

With a full 10 day WU-cache on all rigs!
Berkeley - we are good to go ;-)

119) 留言板 : Technical News : Chaos at the Greasy Spoon (May 24 2007) (消息 575422)
发表于:25 May 2007 作者: Kim Vater
Post:
Hi,

You're not alone on this one.

There's not much activity on the SETI network as can be seen here on this graph:
http://fragment1.berkeley.edu/newcricket/grapher.cgi?target=%2Frouter-interfaces%2Finr-250%2Fgigabitethernet2_3;view=Octets;ranges=d

It's early morning i California - so they will look into the problem soon. ;-)

Kiva
120) 留言板 : Number crunching : Monitor refresh rate settings........ (消息 575293)
发表于:25 May 2007 作者: Kim Vater
Post:
HI,

It could be a registerproblem as leftover from old gfx-drivers.
F.ex if you earlier had a CRT running at 100Hz - the driver then somehow falls back to that setting.
There's "Drivercleaners" out there that completely removes leftovers from previous driver installs. Google is your friend.

Alternativly you could look for another KVM switch that can handle reboots without having that PC set as 'active' on the KVM.

I have a cheap 4-port KVM "EMINE ExorView ME401". By connecting an optional poweradapter it can emulate KVM on all channels. It cost (here in Norway) around 60$ with cables, but without the optional poweradapter.

Edit\\ Link to KVM that I Have: http://www.emine.com.tw/product/EM-ME-1.htm
KVM switches has become quite 'intelligent' over the years compared to the analoge switches that I used 10 yrs ago ;-)

Kiva


前 20 · 后面 20


 
©2020 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.