I'm with the 64 bit 6.10.43
Seems to appear in ngXXX, different nr's, OC should be stable but you never know.
Printable View
I'm with the 64 bit 6.10.43
Seems to appear in ngXXX, different nr's, OC should be stable but you never know.
Dr. Dave does "House calls"
PM me a number and I'll walk you through all of it..:D
Which project is ngXXX?
im using 6.10.43 (x64)
and the errors only appear on the ngXXX units
ng872_ 00099_ 18--
ng846_ 00106_ 14--
ng798_ 00075_ 11--
ng777_ 00089_ 5--
and im not OC'ing at all ... :shrug:
this is 1 of the 4 errors i got
<core_client_version>6.10.43</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
ERROR:: Exit at: .\dock_structure.cc line:401
</stderr_txt>
]]>
I am guessing it's the Human Proteome project. Go to the last page of this WCG forum thread and you will find people with similar errors. My advice is to avoid the Human Proteome project.
Go to your Device Profiles page at WCG and uncheck Human Proteome. See if that gets rid of the errors.
Yes, those WU series belong to HPF2, thay have never found the bug with the newer OSes, especially 64 bit.
I'm using the x64 Beta BOINC client, running the Human Proteome Folding - Phase 2 project and not a single error in the log, yet. Maybe they are making progress on a fix with the Beta?Quote:
Originally Posted by Snow Crash
I don't mean to be brash but it is time to celebrate tonight :woot::woot:
Just looked at my BOINC stats and have over 10,000 BOINC points since joining the cause on 4/14/10.
Running my i7-920 at 4000 mHz, 1.3v; keeps the office nice and warm :) Also started running my NAS box's E8400.
Have I said how much I like the WCG client! To be able to run the client at 100%x8 and still be able to do most things on my computer amazes me.
John :D
Thank you John..Quote:
Originally Posted by Johnmark
The Boinc client has come a long way since we started with it back in 2006.
Back then this wouldn't have been possible as it crashed,etc
Today once you get it "tuned" in it's pretty bulletproof.
I have 3 machines downstairs in my office that I never check or have need to and one has been there crunching since December 2006.
A dual Harpertown that is now watercooled..
The other two are the same, set it and forget it.
Go in and vac the fans every 6 months or so.
Thanks again for your help.:up:
Oh, if it was mine I sure would :D But it's intended as a server in a small company I work for. It won't be OCed there, as companies don't like OCing their production servers, right? Unless it proves to be rock-solid in these WCG tests, than maybe I'll leave it OCed.. it shouldn't peak this much in normal server usage anyway ;)Quote:
Originally Posted by Snow Crash
Thanks, I'm presuming I select either "Error" or "Invalid". But, while I have zero "Invalid" projects.. I have same problems like people below:Quote:
You can check for non-BSOD errors by going to the WCG website, login to your account and then change the Result Status filter.
http://www.worldcommunitygrid.org/ms...Time&pageNum=1
Quote:
Originally Posted by Mydog
Yeah, they are all ng*** related as well, with CPU time mostly around 0.01-0.03. And it happened on both OC and non-OC computer(s). Win 7 x64, and Win Server 2008 R2 x64, with BOINC x64 "stable" client v6.10.43 on both computers.Quote:
Originally Posted by Makavre
Done! Let's see if it helps..Quote:
Originally Posted by sierra_bound
Btw, do "In progress" results include those that are finished on my computers, but still not finished on their end? As I've got 3 pages of "In progress".. I mean, I'm assuming it's one result per core, so I should have 10 "in progress" results at most by that logic.. And I've got 3 more pages of "pending validation".. And I've got one "Inconclusive".. whatever that should be :D But it's not from i7 machine, but from my "home PC".. so I'm not worried :)
It seems that they have fixed the issue in beta: https://secure.worldcommunitygrid.or...d,28870#275715 .. Glad it works for you, but I'll wait for it to trickle down to non-beta projects :)Quote:
Originally Posted by Johnmark
Congratz on the first 10k ;) I'm new as well, and it does make me happy to see daily points and work hours steadily increasing on my profile over the last 3 days :DQuote:
I don't mean to be brash but it is time to celebrate tonight :woot::woot:
Just looked at my BOINC stats and have over 10,000 BOINC points since joining the cause on 4/14/10.
Running my i7-920 at 4000 mHz, 1.3v; keeps the office nice and warm :) Also started running my NAS box's E8400.
Have I said how much I like the WCG client! To be able to run the client at 100%x8 and still be able to do most things on my computer amazes me.
John :D
As for the client - seems that everyone likes it, it really is amazing! I wish that more apps were made this effective in load balancing.. Like some video encoders for example :p
just started up my e6400 machine to help crack some skulls :rolleyes:
come on team, only 2.6 billion more pts til we rank up!
LOL! :rofl: Yeah, with this speed we'll catch them just before New Year :up:Quote:
Originally Posted by masterg
in the past 3 weeks # 2 team has been hovering from 12-15000 results per day, in our past 3 weeks we went from 10,000 to 20,000 per day and an average of 513 points per result. do that math >:P
do they have wcg for ps3?
Congratz on the first 10k I'm new as well, and it does make me happy to see daily points and work hours steadily increasing on my profile over the last 3 days
As for the client - seems that everyone likes it, it really is amazing! I wish that more apps were made this effective in load balancing.. Like some video encoders for example :p
Thanks, I'm finally learning how it all comes together and it's an excellent way to contribute/donate our resources to good causes.
And yes wouldn't that be a treat if all programs behaved the way BOINC Manager does! It seems to share resources almost seamlessly, can't say enough about it!!
John
OK, I agree.. my fault.. It will take us 2 years to catch up :) Unless you plan on keeping the growth rate at same speed, but that's not realistic, as XS has finite number of members, and most will join the cause in first weeks after receiving the message.. Ok, some will join on the date when "competition" starts, etc, etc.. but considering all variables, I think it's still realistic we'll catch the other team by the end of this year. Now, you can do the calculations as well and prove me wrong, I won't mind :)Quote:
Originally Posted by masterg
No, sorry.Quote:
Originally Posted by masterg
The first post in the following thread tells you everything you ever wanted to know (and more) about what those "Status" values stand for ...
http://www.worldcommunitygrid.org/fo...ead=6105#86841
couple of quick ways to get people directly into the team is to have them follow the links below.
If you already participate in the World Community Grid you can join the XtremeSystems team
by clicking the web address below and then click on the "Join Now" button on the page that appears.
http://www.worldcommunitygrid.org/te...mId=69B8364XP1
If you have not yet joined World Community Grid, then click on the web address below.
When you register, the team will be automatically selected for you.
You will still need to download and install the BOINC software.
http://www.worldcommunitygrid.org/re...mID=69B8364XP1
Anyway to start old projects ( 3 days old) that appear to be in limbo?
I've got some projects from the date I signed up that are shown as being processed but only new work is being processed and these are languishing in the que or are ?? I've searched WCG on the subject to no avail.
Thanks,
John
WU tend to be done by date sent UNLESS the project has some"HOT" ones they need done.Quote:
Originally Posted by Johnmark
If you want to force those old ones to be done first you can try suspending the ones that are working and then the others should start up.
If the work units are completed, you can go to the Projects tab on the BOINC Manager and hit Update.
The work units with the nearest deadline are always processed first. If the work units in question have not been processed and are just sitting there, you can always delete them. Generally you're given up to a week to process work units after they're sent out.
Thanks for the replies.
The work units are showing up in BOINC stats only as "in progress". If I suspend any given unit in BOINC manager then the next one just starts up. None of the old ones even show up in managers list, so I can't delete them, unless there is some way of deleting them using BOINC Stats?.
Maybe I'm just crunching hot ones, as all I seem to get to crunch are the one taking 5+ hours to complete?
BTW do any of the projects create more points than others? Currently all projects are selected and it seems my daily points are dropping off :(
John
Try just running the HCC program for 3-4 days and see what happens.Quote:
Originally Posted by Johnmark
As to the existing WU, ignore and let the program work on it's own.
One of the great points of BOINC is that it's pretty smart and works well.
Like I said earlier, got machines I haven't touched in a year and it just keeps humming along.
Johnmark ... what do you have your cache set to and are the WUs that are running say "Running high priority"? Normally 3 days cache is more than enough, no need to get big. WCG credits based on the amount of work, not a fixed value per WU. HCMD2 is currently pointing a little higher than the other project (which are all almost the same) but we are talking about 1 point per hour difference, not really worth changing anything unless there are specific projects you are interested in (or ones you are not). If your points are dropping off, I would suggest you check you "Results Status" on the WCG website to make sure you are not throwing errors.
The cache in BOINC manager is set to .5 days.Quote:
Originally Posted by Snow Crash
Don't see anything anywhere about "high priority"? In BOINC manager(x64 beta) the status of WU shows "running".
Checked the log on WCG and there are a few pages where the status is "detached", but all of those are a couple days old. I THINK most of those detached errors are a result of me changing clients and playing around trying to let BOINC stats control things. I've since turned of AMS and using the setting in WCM. Don't know what else can cause "detached" error?
John
High priority work units appear only occasionally when the deadline is imminent. BOINC always crunches the work units with the nearest deadline first.
Hold on a sec ... you mean you see them on the WCG website but not in BOINC Manager? If so then don't worry, they are hanging from when you were having *detaching* issues and WCG will resend them to someone else eventually (when they pass their due date).
I'd like to join the XS team to help you achieve the goal of this project, but I can't keep my PC active for 24/7, only 8-12 hours per day. And my PC is not very powerful. Should I join or not?
The short answer is "everything helps" so feel free to jump in.:up:Quote:
Originally Posted by @lex
That's what I thought, just wanted to make sure. I'm going to read some FAQs and manuals then :)Quote:
Originally Posted by Movieman
One little question, I just came across.
When a WU finishs, BOINC loads it up (as I can see in the Transfer-Window) but marks it as "ready to report". Sense?
yes, and BOINC defaults sending back every 6 hours..Your fine.:up:Quote:
Originally Posted by soya_crack
OR: hit "update" on the left in Boinc manager screen and it will return it in app 1 minute or less
:wave: Hi! I've joined the team and have put a q6600 at 3.4ghz online. A i7920 and a few C2D may also join in the coming days! Every little helps! :)
Yeah I know that but, the thing which confuses me is that BOINC does upload the WU immediately and the only thing it does every 6 hours is to say "Hey Server, I uploaded some WUs." ?Quote:
Originally Posted by Movieman
Well, that's at least how I understand it from my observations.
My understanding was that completed works are uploaded as soon as possible and then every 6 hours WCG validates the pieces and gives people points, but I may be wrong.Quote:
Originally Posted by soya_crack
The WU's are uploaded but usually not taken in by the servers right away. The reason for this is because the servers would be overwhelmed if all completed work units were sent in at the same time. BOINC uses a scheduler which is a program that determines when WU's need to be downloaded to fill your work buffer, as well as when WU's are taken in by the servers. Your network connection to the servers opens up at specific intervals during the day. You can adjust the internal in your Device Profile section at WCG. You can also adjust how big the work buffer is. I think three days is more than enough.Quote:
Originally Posted by soya_crack
After a work unit is completed and sent back to the WCG servers, it has to be validated. Basically at least two identical work units are sent out to different computers, and after they're completed and sent back, a validation program compares them. If the program determines both work units are valid, it awards points to both computers. If one WU is determined to be invalid, another identical work unit is sent out to a different computer. The validation process usually takes anywhere from 1-7 days.Quote:
Originally Posted by Gonzalo-Gonads
WCG updates individual stats twice a day. Though today there was no early update. Must be a problem with the servers.
k , joined .
hope my X6 arrives soon :rolleyes:
I guess cpu power is most important in these things ?
Yes, CPU power is the most important thing.
Can a radeon 5850 be used for WCG ? if so , where can i download that client .
Also , i see the IBM logo on the wcg site a lot . But there is no client for AS400 (iseries) machines ?
WCG does not use GPU power. That may change in the future.
IBM is the main sponsor of WCG. There is only one client available for WCG - BOINC. There are versions for Windows 32-bit & 64-bit, as well as Linux and Mac OS X.
http://boinc.berkeley.edu/download_all.php
Just added a Q6600 last night, and will probably add a P4 (3.0e) today
(kinda debating if i should prep and add my laptop,.... ) :p:
Quote:
Originally Posted by Makavre
Yes :up:
Laptops require additional consideration for cooling. Plenty of people crunch hard with them, but you have to make sure they stay cool or they won't survive.
If you are never going to use you laptop again, what you can do is take off all the keyboard panel. Also switch the thermal paste. It dropped my cpu by about 40c with the fan hole open and some as5 on the cpu and gpu.
my main system is water cooled so i know a few things about cooling, :p:
but on laptops im completely lost .... :shrug:
would one of these be enough ??
http://images.frys.com/art/product/b...076608.big.jpg
(tearing it apart is not an option right now, .... as i sometimes use it for presentations and demos with clients)
any suggestions will be greatly appreciated :)
These days we can do waaaaay better than AS5 (search for the threads on IC Diamond :up: )Quote:
Originally Posted by snoro
Yes, running it on something like that would be a great help :up: Laptop cooling isn't significantly different to any other air cooling setup in principle ... except everything is smaller and the system will therefore run hotter. Giving it better airflow, keeping it off insulating surfaces, keeping the lid open etc all help :up:Quote:
Originally Posted by Makavre
The two crunching laptops I have I just turn them upside side so the intake fan can breathe and oh boy does it make a difference about 20C worth.
Two questions:
1.) If I tell the BOINC client to update, why don't I see the tasks that were "Waiting to Report" show up in "My Grid"?
2.) I see that at the top of "My Grid" I have done a whopping 8 results, but only 7 went to the team. Is this normal?
Thanks!
The systems we're interfacing with, the WCG servers, have a HUGE amount of results coming in 24 hours a day. As such your personal stats are not "real time" but are updated every 6 hours (I think) with the team stats being updated every 24. If that result you uploaded doesn't show up instantly, don't worry, it'll be there soon enough.Quote:
Originally Posted by retro77
As for the second point, you may well have returned a result before joining the team. At WCG, anything you do with one team STAYS with that team. Conversely that also means that if you return work for any other team, or no team at all, you can't move it to your new team. That's just how it is.
Did you mean upside down? You must have it closed to do that, right?Quote:
Originally Posted by =[PULSAR]=
My i7-720QM reaches 78C while crunching. That's pretty hot.
Thanks D A! They showed up.
edit: I'm up to 17! woohoo.
Yeah sorry I meant upside side down, diarrhea of the fingers haha.Quote:
Originally Posted by xenolith
thats the only thermal paste i had lying around so i didnt had much choice and local computer store dont even have as5 so good luck getting some mx-3 ic diamond or shin-etsu over here at a computer shop.Quote:
Originally Posted by D_A
This gave me an idea so I just rigged up something. I put a 3"x5" cardboard box and a small fan blowing under my laptop and it lowered the CPU temp from 78C to 67C. That should make it safer to run 24/7... I hope.Quote:
Originally Posted by =[PULSAR]=
Thanks for the food for thought. :D
I have the same problem (Central New South Wales, Australia) so I get my nice toys via "mail order" :up:Quote:
Originally Posted by snoro
There's places that will e-tail some ICD and post it to you, I have no doubt.
yeah i can get some shipped here but i didnt wanted to make a order just for some thermal paste for my notebook if as5 would be quite enough. Temps are now lower than 60c under 24/7 wcg load. Would be useless to ask for more.Quote:
Originally Posted by D_A
Once I exhaust my supply of ceramique I'll be replacing it with ICD, since I've found an e-tailer here in Auz ... might be a while though.Quote:
Originally Posted by snoro
me its next time i buy either my cpu watercooling block or when i will buy heatsink for my dedicated cruncher rig ( which i still dont know what it will be made of)
Need a little help here. Tried looking for an alternate to BOINC view, to try and manage 2 friends rigs, along with my 2 here at home before the May 1st Super Computer project. Boincview seems to be down. Could someone give me at least a push in the right direction, if not a good step by step? Thank you.
Tips on installing BOINCView
I believe the BOINCView site is dead. But there are a couple other places where you can download it. Here's one. Link
So regarding laptops, bottom line is - if you've got a laptop that's having warranty running out in a month or so, best way to get a new one is running BOINC client on it 24/7.. or as long as it lasts :D
(just kidding, don't even own a laptop :) )
I've been running my laptop 24/7 since they day i got it. 100% with no added cooling. The 1 year warranty is almost over and no sign of me getting a new laptop... idk what gives. :shrug:Quote:
Originally Posted by LuxZg
:rofl:
Hey, I just joined the team. Hopefully we can reach the goal! I'll make sure my friends will start folding as well ;) Too bad that vulcano erupted. I've ordered a W3530 and DDR3 RAM, but both won't arrive untill the 5th of may. Prolly because of the fact the airspace was closed for quite a while :( I will use my Q9650 @ 4050mhz to it's full potential though! :up: Anyway, a few questions in getting the maximum amount of PPD. They might have been answered before, but I currently do not have the time to read through the entire thread. I might be able to do that this weekend.
Stuff I did so far:
-Switched to advanced view in Boinc
-Set "While processor usage is less than" to 0, so it's unrestricted.
-Set "Use at most ... % CPU time" to 100.
Anything else I'm missing in maximizing my PPD? Also, how can I check whether I'm running the 64bit or 32bit client? Thanks! :up:
If I didn't need my laptop I"d be crunching on it, and it will fail XD Moderate load and cores hit 85C, even smelled something burning once :rofl: Pretty sure they messed something up when the put it back together after the MB failed right after I got it.
Try covering it with a blanket.. my ex did that, and sure nailed that hard drive from the first try :rofl: I bet it would be a fire if it was doing anything else except just showing a web page she was looking at when she fell asleep :DQuote:
Originally Posted by Otis11
Anyway, I wanted to brag a bit - I've got TWO BADGES! :)Didn't even know these existed, and today I went to see my stats and if there were any errors and such, and yup - there they were :) Sure, it's just bronze, but NOT BAD! :)
As a side note, no more "errors", but I had to discard all projects except the Childhood cancer.. I'm guessing now I'll have to add them one by one to see which ones show those weird errors.. ng*** projects was obviously that HPFolding-p2, but than nh*** errors occured, so I just removed all projects except one. Helping children seemed nice, so I've left that one. No errors.. Now, if anyone know which data parts are tagged with nh*** names I could try activating several at once.. so please tell if you know. Cheers everyone!
TWO BADGES!!! :D
EDIT: Btw, how come I've got like... ~70 days of runtime when I was running it for what.. 9 days? Is that 9 days times number of cores? Like i7 has 4 cores (but 8 threads), and my Core2Duo has 2 cores (2 threads), which is total of 10 threads.. Or how does it count actually? Counting it as hours-per-thread is only way I could get to such a high runtime number in under 10 days (plus i7 was at least a day late behind my C2D..explaining lack of "some" runtime hours..). Weird.. but it does make a person feel better :)
I started in June '09 and currently am at around 9+ years with a couple of core 2 quads so yes your runtime numbers make sense and your explanation is good
My Vista rigs don't like one of the projects but I don't remember which so I put them on cancer and childhood cancer and all is just fine
I think the runtime is calculated based off how long it took to complete each WU you returned. Dividing the listed runtime by number of threads should result in a more accurate runtime I think.
Also for me it was the Human Proteome Folding - Phase 2 project that almost always generated an error early in the computations.
---
Is the server down? I haven't been getting new work units and I'm down to a few hours worth of work :(
Edit: It just waited until it had completed every work unit before getting new ones, bit of a waste for mutli-threaded systems :\
question about boincstats, how do I make my Number Of Hosts (Active Hosts) not hidden?
Human Proteome Folding - Phase 2 is the only project I ever really get errors on. It has had problems on Vista and Win7 machines, so that is probably your problem. :up:
And as far as the run time - it's hours each thread gets to run. So if you don't touch your i7 you'll get 8 days worth of credit in 24 hours, but it takes some time to validate, so it may be delayed a bit.
And idk where to change the host thingy, sorry. :shrug:
I think I found out actually. But it wasn't on Boincstats settings, I was looking in wcg website on my settings there and saw "show hosts" hereQuote:
Originally Posted by werdwerdus
Well I managed to add 200MHz to my overclock (3.8 to 4.0 on the E6300). Upgraded BOINC to the 64-bit version (6.10.43), but now my GPU (Radeon HD 3850-256) isn't being identified.
22 Apr 2010 2:42:54 AM Starting BOINC client version 6.10.43 for windows_x86_64
22 Apr 2010 2:42:54 AM log flags: file_xfer, sched_ops, task
22 Apr 2010 2:42:54 AM Libraries: libcurl/7.19.7 OpenSSL/0.9.8l zlib/1.2.3
22 Apr 2010 2:42:54 AM Running as a daemon
22 Apr 2010 2:42:54 AM Data directory: C:\ProgramData\BOINC
22 Apr 2010 2:42:54 AM Running under account boinc_master
22 Apr 2010 2:42:54 AM Processor: 2 GenuineIntel Pentium(R) Dual-Core CPU E6300 @ 2.80GHz [Family 6 Model 23 Stepping 10]
22 Apr 2010 2:42:54 AM Processor: 2.00 MB cache
22 Apr 2010 2:42:54 AM Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 syscall nx lm vmx tm2 pbe
22 Apr 2010 2:42:54 AM OS: Microsoft Windows 7: x64 Edition, (06.01.7600.00)
22 Apr 2010 2:42:54 AM Memory: 6.00 GB physical, 6.00 GB virtual
22 Apr 2010 2:42:54 AM Disk: 298.09 GB total, 119.90 GB free
22 Apr 2010 2:42:54 AM Local time is UTC -4 hours
22 Apr 2010 2:42:54 AM No usable GPUs found
Any way to get MilkyWay@home-GPU to work again? I tried updating from Catalyst 10.1 to 10.3 but that didn't work. I'm trying to switch back to 9.12 (iirc) now, partly because CCC really sucks. MW@h-GPU was working pretty well before, and I regularly got >18k BOINC ppd from it.
BOINC doesn't do GPUs....
edit: sorry WCG doesn't do GPUs....
I know. But I have an ATI (RV670) GPU, and Milkyway@home supports RV670's. I had it running pretty damn well before I updated to the 64-bit client.Quote:
Originally Posted by retro77
http://img9.imageshack.us/img9/2189/69340444.jpg
My IT side says go back to the 32bit client, unless the 64bit client, without the GPU, can out perform your 32bit client with the GPU.
Are you running 32bit or 64bit drivers?Quote:
Originally Posted by Bobsama
64-bit of course. The same as before. What was odd was swapping to .43_64 seemed to be what broke GPU side of it. I'm going to try reinstalling tomorrow.
And yes, the 64-bit client is worthwhile. It's about 10% more PPD than 32-bit, and I'm not sure that going back to the old version would restore GPU.
Odd part is though that even after installing the 64-bit client I'm seeing 32-bit processes...
Nope, that's normal. The work units are all 32bit. The 64bit client is just slightly more efficient with memory and process allocation.
Now that my rig is doing WCG now anyways, I started looking into utilizing my GPU (5870) as well. I decided to start using Milkyway@Home via BOINC, but a problem came up: It's automaticly using both the CPU and GPU client, causing my CPU to work for Milkyway@Home as well. I don't mind helping them out, but fighting diseases is far more important in my opinion, so I'd like Milkyway@Home only to use the GPU client. The question is: How? Thanks! :up:
Go to the Milkway website, log on to your account, select Milkyway Preferences, Edit preferences, uncheck the CPU box. :yepp:Quote:
Originally Posted by Musho
Thanks! That fixed it! :clap: Now let's continue curing those nasty diseases :yepp:Quote:
Originally Posted by Snow Crash
And if you don't care about poor performance with Aero, you can turn on the use GPU while in computer in use in BOINC, and enable it 24/7.Quote:
Originally Posted by Musho
... or you could just turn aero off
Hi all,
Been looking around the forums and noticed the "Lets build a super computer" announcement and decided I wanted to get on board.
Ive been F@H on and off using the gpu's in my sig and just installed WCG today.
I followed some of the threads here and got it setup and joined the team.
So far the rig in sig is moving along at full load and Im about to get a few more rigs going as dedicated.
Those being a AthlonIIx4 620@3.4ghz, Athlon64x2 4450e, and a Athlon64 3200 @2.5ghz.
I guess my question is if there are any tips I should know to improve my output?
Does overclocking the cpu give any gains?
And is there any hardware that may not be up to the task?
I appreciate the input in advance.
Slave
Welcome slaveondope!
Just run the client
Overclocking helps
I'm running it on a netbook Intel Atom N270...
Anything from a PIII up can help. As for improving performance there's multiple threads (probably should consolidate them all ... again) on the topic. Just keep reading and you'll learn fast. :up:Quote:
Originally Posted by slaveondope
Figured it out. Before, it was installed as a Service. Apparently, running as a service in NT5 would have worked, but running it in NT6 doesn't. Alright. I'll lose a few ppd to running MW@h but oh well.
Trying to get Milkyway running the GPU only and letting the CPU crunch for WCG and having problems. MW is not getting any work? And I'm not sure what I should be doing with CPU setting?
Here are screen shots of milky way config. pages.
http://i168.photobucket.com/albums/u...07/MW_pref.jpg
http://i168.photobucket.com/albums/u...references.jpg
When I ask BOINC Manager to update MW, the log shows "not reporting or requesting any task
Any ideas are appreciated.
John
One question.
I had to clean install the OS, but my BIONIC and BIONIC Data-folders are on another drive and still in tact. Can I just reinstall the BIONIC-client on top of that, and what happens to my running/unfinished tasks?
There's a good chance it will just pick up where it left off.
Thx, it worked :)Quote:
Originally Posted by D_A
Just installed it on top of the old one, and it piked up my old tasks :), just like nothing has happed.
I had the same problem at first. Turns out it's not using the GPUs when the computer is already in use. To fix that, in your BOINC client click "Advanced" and then "Preferences". Go to "Processor Usage" and tick the box saying "Use GPU while computer is in use". That fixed it for me. The windows UI is slightly laggy, but nothing serious and I'm able to use my PC for anything besides gaming just fine.Quote:
Originally Posted by Johnmark
I also set "While processor usage is less than ... percent" to 0, so it's unrestricted. Should boost the PPD for the WCG project. I also set "Use at most ... % CPU time" to 100. Good luck! :up:
Edit: What's the PPD goal for the 1-8 may week? The announcement says Movieman wants to try and beat 30k-40k P4s, but I'm unsure how much PPD that would equal.
At a guess ... 50 million. Quite doable if every member of the greater forum community threw one system on the project.
Thanks Musho,Quote:
Originally Posted by Musho
I already did that and no go. I've since disconnected that project as I don't want to mess up things before the May1. I'll work on it later in May.
John
Got home from work today to find that not all work units all not all being sent out? Had 28 units sitting there waiting to be sent but yet BOINC Manager had sent just one, and that was minutes ago? It just left the finished units sitting there?
I was almost out of work so changed the amount of work to one day's worth and poof it sent out all the finished units and downloaded more? Is this normal ?
P.S. D A , how are you able to display your badges, and what is a "stone" which I heard referenced in another thread?
TIA
John
@ Johnmark
For posting badges, Ive just been copying the image URL from the WCG website as I get them and pasting it into my sig. If you've been checking your stats on wcg website they will show up there as you earn them.
'Stones, unless I'm mistaken, usually refers to point milestones for WCG and BOINC. They are posted in this thread http://www.xtremesystems.org/forums/...d.php?t=189054. For example, tomorrow I'm going to be hitting my first million BOINC points. Not that I'm looking forward to it or anything...
I've noticed I've had to manually tell mine to report. Global issue?
The work units? Usually it only contacts the scheduler to report tasks and get new ones a couple times a day if that's what you're asking. If you want to get BOINC to report WU's immediatly on its own, I have a tutorial on poppageeks' website. http://www.poppageek.com/howtos 3rd one down.