Message boards :
Number crunching :
Are my whetstones and drhystones low for my machine??
Message board moderation
Author | Message |
---|---|
eberndl Send message Joined: 12 Oct 01 Posts: 539 Credit: 619,111 RAC: 3 |
My computer just did its benchmark. 586 double precision MIPS (Whetstone) per CPU 1475 integer MIPS (Dhrystone) per CPU I have a 1.6 GHz machine running XP. I hadTrillian (an IM program) as the only other program open. It takes me about 6 h/WU. Is this high or low? Is there a way to speed up my units (w/o over claocking). <img> <br /> Feel free to take <a>a look inside my brain[/url] |
Petit Soleil Send message Joined: 17 Feb 03 Posts: 1497 Credit: 70,934 RAC: 0 |
Hi eberndl, It seem a little bit low indeed. I guess it's your Laptop ? Mine's benchmarks are ; 711 double precision MIPS (Whetstone) per CPU 1826 integer MIPS (Dhrystone) per CPU Presario 2800 - P4m 1.4 GHz - 400 fsb - 512 Mb DDR 266 It also takes 6 hours per WU Here's a little trick to score higher benchmarks. Restart you machine, do the benchmarks, and the benchmarks again. Results will be higher the second time. EDIT // Overclocking your Laptop isn't a good idea. Amicalement Marc |
eberndl Send message Joined: 12 Oct 01 Posts: 539 Credit: 619,111 RAC: 3 |
But should I' be going a _little_ faster than you? and my last two units were 6.5 h, and that number has been creeping up... are the units longer or should i do a disk defrag or some such? <br /> Feel free to take a look inside my brain |
Petit Soleil Send message Joined: 17 Feb 03 Posts: 1497 Credit: 70,934 RAC: 0 |
Defrag won't help. SETI doesn't use the disk much during processing. Did you check the task manager ? SETI should be running at 99%. EDIT // What are your machine specs ? CPU - RAM - etc |
Petit Soleil Send message Joined: 17 Feb 03 Posts: 1497 Credit: 70,934 RAC: 0 |
eberndl ?! eeeeeeeeeeeeeeeeeberrrrrrrrrrrrrrrrrrndl (with echoes too) I hope her PC is still working. Or perhaps she've been abducted by alliens!, eaten by Grizzou!, doing Loundry!, having dinner!??? Stay with us ! as the search for eberndl continue, right afer these messages. |
Heaphus Send message Joined: 1 Apr 03 Posts: 96 Credit: 4,148,549 RAC: 0 |
Keep in mind also that benchmarks come in lower on v4.09 than they did with v4.05. In other words, make sure your comparing apples to apples. And, that's not to mention the whole host of other reasons one computer will be faster or slower than another. |
Petit Soleil Send message Joined: 17 Feb 03 Posts: 1497 Credit: 70,934 RAC: 0 |
> Keep in mind also that benchmarks come in lower on v4.09 than they did with > v4.05. In other words, make sure your comparing apples to apples. And, that's > not to mention the whole host of other reasons one computer will be faster or > slower than another. Absolutely. But she did mentioned that crunnching time "has been creeping up" It's true that the benchmarks are lower with v4.09 but real crunching time never changed on my side. It's always 6 hours +/- 10 minutes. I'll start to worry too when it slowly start to take 6:15, 6:30, 6:45 and so on. -.-. --.- -.. -..- . - --... ...-- .-.-. -.- |
ChinookFoehn Send message Joined: 18 Apr 02 Posts: 462 Credit: 24,039 RAC: 0 |
|
Petit Soleil Send message Joined: 17 Feb 03 Posts: 1497 Credit: 70,934 RAC: 0 |
> Since Seti is a minor project for me (set at just under 3% processing > allocation); I do not pay too much attention to it. I have noticed that > some of the recent work units, are taking around ½ an hour while others, over > 1 hour. That's strange. I have processed about 200 WU on my PC Laptop and about 60 on my MAC since June 22nd and they all took 6 and 10 hours respectively to proceed. Now I am currious about the machine your using to crunch one SETI unit in 30 minutes ? > However, I did read on the LHC site (before the massive cull of > threads), from one of the administrators, that his 'crunching time' was > taking longe; until he defragmented his drive - even though Windows stated it > was unnecessary. I realy don't think fragmentation play a big role in processing time. It might have been a coincidence. The file is what 350K ? the results 16K ? Pretty hard to frag a 16 Kb file I guess. > It is something I do regularly and the only thing I have noticed is that my > Phase 3 CPDN time-steps are being calculated faster than the Phase 2 > time-steps were which were calculated faster than the Phase 1 time-steps > (which may be a CPDN process) - especially for my #1 cpu on my P4-HT. I really don't know as for CPDN since I am not running it. |
ChinookFoehn Send message Joined: 18 Apr 02 Posts: 462 Credit: 24,039 RAC: 0 |
|
Petit Soleil Send message Joined: 17 Feb 03 Posts: 1497 Credit: 70,934 RAC: 0 |
> The only computer I am running Seti on, at present, is my Athlon 2600+ > (Barton) which is slightly over-clocked with 1GB RAM. The latest work > unit, when completed (which might not be for another day or two), will > likely take around 1hr 20min. > > -Richard Pretty impressive results for a 2600+ you really made a good over clocking job ! |
bjacke Send message Joined: 14 Apr 02 Posts: 346 Credit: 13,761 RAC: 0 |
Hi, I'm running an AMD athlon 2800 XP+. I had it overclocked to 2300 MHZ an I needed 2h30min. @H. Richard Utzig: Would you tellme your settings? The whole is more then the sum of its particles. Aristoteles Best wishes from Berlin(52°35'N,13°23'O), Basti S@h Berkeley's Staff Friends Club © member |
eberndl Send message Joined: 12 Oct 01 Posts: 539 Credit: 619,111 RAC: 3 |
ran a disk defrag (no, Grizzou did not eat me, but thanks for the concern), I don't know if it's going to make things faster or not... cause just found out that the whole night my cycles were being eaten by a program I'd already closed (my html editor has a loader that likes to use 70%+ of my cycles) Something also caused my CP unit to be unexpectedly killed... maybe it ice balled, but I don't think so... Anyway, I'll get back to you later today when I have some SETI times on my comp.... in the mean time: work. <br /> Feel free to take a look inside my brain |
Stephen Balch Send message Joined: 20 Apr 00 Posts: 141 Credit: 13,912 RAC: 0 |
eberndl > ... cause just found out > that the whole night my cycles were being eaten by a program I'd already > closed (my html editor has a loader that likes to use 70%+ of my cycles) > Now you've got me curious. What HTML Editor do you use? Cheers, Stephen <P>Not only are we doing distributed processing, now we're doing distributed debugging.<BR>Who'd a' thunk it? (GRIN)<BR><a> |
ChinookFoehn Send message Joined: 18 Apr 02 Posts: 462 Credit: 24,039 RAC: 0 |
|
bjacke Send message Joined: 14 Apr 02 Posts: 346 Credit: 13,761 RAC: 0 |
> Vielleicht liegt es mehr am Mutterbrett und ich benutze ein Shuttle XPC. > > -H. Richard Utzig Well you have got a piont there. That will it be! The whole is more then the sum of its particles. Aristoteles Best wishes from Berlin(52°35'N,13°23'O), Basti S@h Berkeley's Staff Friends Club © member |
Heffed Send message Joined: 19 Mar 02 Posts: 1856 Credit: 40,736 RAC: 0 |
> The only computer I am running Seti on, at present, is my Athlon 2600+ > (Barton) which is slightly over-clocked with 1GB RAM. The latest work > unit, when completed (which might not be for another day or two), will > likely take around 1hr 20min. Well then your machine is outperforming the biggest guns your home computing dollar can buy. Funny that.... |
ChinookFoehn Send message Joined: 18 Apr 02 Posts: 462 Credit: 24,039 RAC: 0 |
|
M.E. Robbins Send message Joined: 4 Dec 01 Posts: 1 Credit: 488,394 RAC: 0 |
I'm with Heffed on this one Richard. I've crunched hundreds on WUs on several chips (some quite a bit faster than that 2600+ of yours), and I've yet to break the 2 hour mark by a whole lot. Keep in mind that the completion bar is a terrible thing to base projections on, as the first 60% streams by pretty quickly while the remaining 40% takes quite a bit longer. It'd be more helpful if you'd provide actual times for completed WUs rather than projections. |
eberndl Send message Joined: 12 Oct 01 Posts: 539 Credit: 619,111 RAC: 3 |
it's HTML Pad 2004 Pro... but a friend of mine gave me a crack so I could use it without paying for the upgrade... but the crack seems to eat away at my cycles if I don't turn it off... which I didn't last night. <br /> Feel free to take a look inside my brain |
©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.