Welcome to MilkyWay@home

Posts by scottishwebcamslive.com

1) Message boards : News : and a little payback (Message 48720)
Posted 11 May 2011 by scottishwebcamslive.com
Post:
Seconded +2!


@ Beyond
Or just take a period of past days and run a script to double the user credit for that period. No extra server load at all :)


nice idea too

Ian
2) Message boards : News : and a little payback (Message 48711)
Posted 11 May 2011 by scottishwebcamslive.com
Post:
hello,

the double credit idea is ok and welcome but wouldnt you be better not announcing when it takes place

1. this will lead people who very rarely if ever crunch for the project suddenly turning up then vanishing after the double credit period is over

2. connected to 1 really is the massive load on the server of streams of crunchers from all over trying to download and upload WU if the server has any weakness at all this extra load will crash the server for the rest of us

believe me ordinary crunchers that work here and deserve the credits will notice within minutes there getting extra credit

just my two pence worth
Ian
3) Message boards : Number crunching : 11.5 & 11.5a Hotfix Released (Message 48690)
Posted 10 May 2011 by scottishwebcamslive.com
Post:
hi.

what temps are your VRM'S sitting at with the app and 775 meg clock

i'm interested because as you know i'm running the stock app but at 800 meg and i'm processing the 267 credit units in 130 seconds only 5 seconds slower than the alternitive app but at the end of each WU as you know theres a little gap which is giving my cards a little break long enough to take the edge of the temps on the VRM's
i cleared some cables out the road of the twin 4 inch fans i have directed at the cards and something strange
the temps of the VRM's stayed at exactly between 100 and 103 degrees depending how far they were through the work unit but the gpu temps dropped by 2 degrees accross the board
just shows its alot harder to get the temps down on those VRM's than the GPU's which is where the trouble starts with these cards :(

i'd be interested to see if the new drivers somehow do take a little bit more off the temps especially the VRM's which are now my only problem as the 4 GPU's sit betwee 84 to 89 degrees peak

best regards
Ian
4) Message boards : Number crunching : if the apps have no benifit why are all the top hosts using them ? (Message 48613)
Posted 8 May 2011 by scottishwebcamslive.com
Post:
hi,

crossfire just allows me to get both cards working without having to make dummy plugs or anything else
always wonderd if theres a scaling differance mind you as crossfiring cards hurts there doubling up performance
i learned that when x-firing 3three 4890's the first card would do around 100,000 credits a day add another and that only went up to just under 180,000 and if you added the third it would pretty much peak at 240,000 instead of the 300,000 you might have thought
so as i say i often wondered if by not x-firing them then they would actually double up but couldnt be bothered doing the testing or making the dummy plugs lol

no i wasnt going to disasembly the card in any way just place the cooler on the black back plate somehow lol

see the problem is that its only one of the cores out of the four thats hitting 90 degrees the others sit around 85 to 87 but as soon as that core tips over the 90 degree mark then the fans become very noisy
and the ambient temp in the room can have a differance as it nudges up a degree so does the card and hay presto your just over 90 degrees again

so really i'm not looking for 10 degrees or anything just keeping that one core down to 85 degrees with the rest would make me happy as the fans although audible are not deafening

the overpressure principle doesnt work with this box mainly because i cant get the side on with the cpu cooler which is huge
plus i've always found aiming multiple fans directly inside the open side of the case kept the temps lower than closing everything up and forcing air through the box

i've been dealing with computers as a hobby for 30 years and have got to the stage now where it doesnt have to look pretty or conform to logic if it defies logic and works thats good enough for me lol

ian
5) Message boards : Number crunching : if the apps have no benifit why are all the top hosts using them ? (Message 48610)
Posted 8 May 2011 by scottishwebcamslive.com
Post:
hi,

thanks but the board your using appears to have way more space between the first 4 blue slots so that i think youd at least have some gap if you put one card in slot 1 and the other in slot 4 youd have a rather nice gap without running out of motherboard lol

the motherboard you mention looks great unfortunately i pretty much only ever upgrade motherboards if they break or some new cpu chip config comes out making the board usless

i was thinking of somehow attatching a cpu cooler somehow to the metal backplate of the first card see if that has any effect on cooling the heatsink
not sure how i'm going to do that yet though

by the way i take it the red openings at the end are intake holes as i was thinking of sticking a fan pointing directly into them too lol

Ian
6) Message boards : Number crunching : if the apps have no benifit why are all the top hosts using them ? (Message 48604)
Posted 8 May 2011 by scottishwebcamslive.com
Post:
wow thanks guys loads of useful info

i'm using the Gigabyte GA-X58A-UD7 Intel X58 motherboard
http://www.overclockers.co.uk/showproduct.php?prodid=MB-229-GI&groupid=701&catid=5&subcat=1692
which when i look at it has 3 blue pcie slots together which just now have each of the cards in slot 1 and 3 the only other blue slot i can see i dont know if its a proper slot for x-fire is a slot in fifth pace the other side of a white pci slot there is a gap of some two or so inches from slot three where the second card now resides and the 4th blue slot that i take it is another pcie slot that can be used for crossfiring these two cards
that would mean the first card in the 1st blue slot and the second card in the 4th blue slot which would need a bridge of about six inces or more so i cant see that being correct
plus the 4th blue slot would squeeze the card against the bottom of the case even though it is a pretty large case or it was but not maybe compared to the monsters people use today

this is still quite an expensive motherboard so would like to think that all of these 4 blue slots can be used for full pcie16 data transfer ?

then i'd just have to get a bigger case as the second card would hang of the end of the motherboard

let me know what you think

Ian
7) Message boards : Number crunching : if the apps have no benifit why are all the top hosts using them ? (Message 48591)
Posted 8 May 2011 by scottishwebcamslive.com
Post:
hi,

just made each core have 2 wu each and as expected slightly more heat running around ( maybe because of more work being done ? )

and as expected the wu have doubled in there crunching length

but i had a look at your system wu and nearly fainted how on earth are you getting work units finishing in around 60 and 90 seconds for each differant type if your running twin wu on one core ?

mine are now running at between 200 and 250 seconds each which is way slower than yours even though my cards are only modestly overclocked i'd have to overclock the thinks to 2 ghz to get anywhere near those times

so is there something i'm missing here ?

Ian
8) Message boards : Number crunching : if the apps have no benifit why are all the top hosts using them ? (Message 48585)
Posted 8 May 2011 by scottishwebcamslive.com
Post:
hi

thanks for your quick reply as i say i tried this a long time ago so maybe theres more of an effect now

I'm running a i7 920 at 4 gig with twin 5970's using 4 GPU's

so any increase in throughput i suppose would be even more amplified

my only problem with overclocking the 5970's is heat and lots of it ( what do you mean unlocking the shaders ? ) i though on this project fiddling with all that sort of stuff made no impact as it was sheer GPU power that was needed

i have to run them at stock and lower mem clocks just to keep the temps around 85 degrees :(

whoever designs the air intake designs on these cards needs shot because when you use two they basically sit slap bang next to each other and even sticking something in between them to seperate them as mush as possible and aiming twin 4 inch fans at the gap doesnt help much grrr

so if running two wu per core makes extra heat then i may not be allowed to do it anyway :(

best regards
Ian
9) Message boards : Number crunching : if the apps have no benifit why are all the top hosts using them ? (Message 48578)
Posted 8 May 2011 by scottishwebcamslive.com
Post:
hello,

i used the app to double the workunits per core ages ago but saw practically no differance in throughput
so i'm wondering why so many of the highest rated hosts on this project are using them ?
is it because of the multiple gpu's per machine there running that there is a benifit to running 8 work units from 4 gpu's ?
or do you get more of a bank of work units downloaded to your machine because it thinks you have 8 cores instead of 4 ?

there must be some reason i just cant fathom it
something else i cant fathom now that i've returned from an extended holiday is that there seems to be two differant types of wu with differant times to complete
now thats fairly ok but what i dont understand is the various differant times the same type of work units seem to be crunching at

thankfully i havent had the failed work unit problem some have had or not that i've noticed but i did like the work units of a same type finishing in a same amount of time as it helped in fault diagnostics etc etc

sorry about all the questions

best regards
Ian
10) Message boards : News : testing new credit policy on nbody workunits (Message 48547)
Posted 7 May 2011 by scottishwebcamslive.com
Post:
hi,

did i miss the double credit day ?

Ian
11) Message boards : Number crunching : 11.4 Catalyst Production Version - Red Flag for 5970s (Message 48542)
Posted 7 May 2011 by scottishwebcamslive.com
Post:
hello,

i have twin 5970's that i just put in my main machine so downloaded the new 11.4 drivers as the machine had none installed
everything was fine until i started boinc with mw on it
it started 4 work units up and was chewing them up then two slowed one almost made it to 98% and the the pc froze
its done that over and over again now so i think i'll try the 11.3 drivers see if they woirk with my twin cards

will be back lol


hopefully with good news

Ian


UPDATE : downloaded and installed the 11.3 drivers and everything so far seems ok
quite amazing how the small change from one driver to the next supposed upgrade can quickly cause all sorts of problems

thanks Zydor for the thread and your fix for the problem

best regards
Ian
12) Message boards : Number crunching : are we back to the normal WU now ? (Message 46295)
Posted 19 Feb 2011 by scottishwebcamslive.com
Post:
hi,

if were back to the standard WU then i can hook up my other 5970 in crossfire again instead of it just sitting their looking pretty :)
as in crossfire they just crash the machine on the longer WU's

best regards
Ian
13) Message boards : Number crunching : We have some new long running WU's!!! (Message 46083)
Posted 9 Feb 2011 by scottishwebcamslive.com
Post:
Hello,

the new work units are hugely resourse hungry so much so one of my machines just crashes as soon as you start milkyway
for some reason the wu are doing the same thing as collatz in that box and using all the 4 gig of ram and loads more with everything else switched off including all other projects
thats on a twin 5970 x-fired machine but even on my main machine only using 1 4890 another 1 gig of ram is being used that wasnt yesterday for some reason by MW

as i say collatz does this as well making it unusable on that machine i just hope MW hasnt gone the same way !

best regards
Ian
14) Message boards : News : a few drive errors (Message 46048)
Posted 7 Feb 2011 by scottishwebcamslive.com
Post:
Hello,

I can assure you neither of my machines has recieved any work for nearly 12 hours now so the servers are not as you say " running again now " !

Ian
15) Message boards : Number crunching : Server outage then my pc stops asking for work (Message 45733)
Posted 24 Jan 2011 by scottishwebcamslive.com
Post:
hi,

maybe i didnt explain the problem well enough
it only takes the server to run out of work for 10 or 12 mins and my machine runs out of work
even if the servers fill up with work after 20 mins my pc will just sit there and not ask for any of the new work
my other machine ( far slower ) will continually run and download work although it very rarely actually runs out of work unless the servers down an hour or more

any ideas without attatching to other projects ?

best regards
Ian
16) Message boards : Number crunching : Server outage then my pc stops asking for work (Message 45725)
Posted 23 Jan 2011 by scottishwebcamslive.com
Post:
Hello,

this is a problem that happens alot and is very annoying
whenever the server hits a glitch or whatever my pc runs out of work minutes later
when work is on the server again my machine will just sit their not asking for any
today it sat for 5 hours without asking for any until i noticed it and just forced it too
is there a way to force your machine too ALWAYS look for new work every few minutes so youve always got the most you can have ?

best regards
Ian
17) Message boards : Number crunching : Validator stuck? (Message 45694)
Posted 22 Jan 2011 by scottishwebcamslive.com
Post:
Hi,

lets see how high it will climb :)

Workunits waiting for validation 105,997

Ian
18) Message boards : Number crunching : Host with WAY too many tasks. (Message 45658)
Posted 21 Jan 2011 by scottishwebcamslive.com
Post:
hello,

how on earth do you convince boinc that any CPU has thousands of cores on it ?
i thought that was written in stone your pc only told the truth about it ?

best regards
Ian
19) Message boards : Number crunching : Aaargh! Servers are out of new work!(2)" (Message 45631)
Posted 19 Jan 2011 by scottishwebcamslive.com
Post:
hello,

there is always a third way of doing this which would involve alot more work on your part i would suspect
that is make the WU way longer than 90 seconds on some of our machines ( say 10 or 15 minute length or longer each ) and let us choose to do those WU on our faster machines if you dont want to figure out some way of sending them out to faster machines automatically

just a thought

Ian
20) Message boards : Number crunching : Aaargh! Servers are out of new work!(2)" (Message 45614)
Posted 18 Jan 2011 by scottishwebcamslive.com
Post:
hello,

the problem of running out of work can be cured in two ways

1. you up the amount of WU per core that we can download ( which for some reason your desperate not to do )

2. hold a far bigger cashe of work units on your own server so that even if it jams for an hour or two we still can download work until it starts producing new work again ( although that wont work either .... just now for instance theres 5000 WU on the server page waiting to be worked on it says and yet nothing for my two machines :( )


GPU technology is making leaps and bounds ( twin dual GPU 5970's here ) and having 10 or 15 mins of work when theres a server glitch which can last hours just isnt cutting it anymore

best regards
Ian


Next 20

©2024 Astroinformatics Group