How to disable Global Correlations tasks?

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5872
Credit: 117471437235
RAC: 35503795

RE: My Athlon X2 4800

Message 98149 in response to message 98148

Quote:
My Athlon X2 4800 crashes every GC task it gets.


Well not quite every one. There is a successfully completed GC1 task on the first page of your results list for that machine.

I have a very similar CPU, Athlon 64 X2 4600+ which is overclocked somewhat. It crunches GC1 tasks in pretty much the same time as yours so I imagine yours is overclocked as well. Now that you have hotter weather perhaps you need to reduce the overclocking a bit. The other thing would be to check your heat sink and fan for dust blocking the fins or for any reduction in fan efficiency. Are you monitoring your CPU temperatures?

Cheers,
Gary.

Mark Henderson
Mark Henderson
Joined: 19 Feb 05
Posts: 34
Credit: 38994719
RAC: 3930

It is overclocked to 2.6 from

It is overclocked to 2.6 from stock 2.2 I will reduce the overclock to stock and try it out. My temps are about 50c which I assume is ok.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5872
Credit: 117471437235
RAC: 35503795

Mine is overclocked to 2640

Message 98151 in response to message 98150

Mine is overclocked to 2640 MHz (12x220). I don't run it that fast in summer as it can tend to lockup or trash tasks in the hot weather. I was running it at 12x214 last summer and increased it about a month ago when winter arrived.

Your temps are fine and you don't need to go right back to stock. I'd try it at around 2.5GHz where it should be OK.

Cheers,
Gary.

Mark Henderson
Mark Henderson
Joined: 19 Feb 05
Posts: 34
Credit: 38994719
RAC: 3930

Gary, Reducing the overckock

Gary,
Reducing the overckock from 2.6 to 2.5 seems to have solved the problem of WU crashing on the GC tasks.
This x2 4800 has been crunching 24/7 for over 5 years, I suppose any weak links will begin to show after that much abuse.
Thanks again

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3562358667
RAC: 0

I'm surprised you managed to

Message 98153 in response to message 98152

I'm surprised you managed to go that long. My 939 mobos universally began to falter after about 18-24mo of overclocking.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5872
Credit: 117471437235
RAC: 35503795

RE: Reducing the overckock

Message 98154 in response to message 98152

Quote:
Reducing the overckock from 2.6 to 2.5 seems to have solved the problem of WU crashing on the GC tasks.


I'm glad that you got that sorted out.

Quote:
This x2 4800 has been crunching 24/7 for over 5 years,


Mine's been going for a while too - perhaps a bit less than that.

Quote:
I suppose any weak links will begin to show after that much abuse.


The overclock isn't excessive and I very much doubt that the overclock itself is any problem at all. With the experience of hundreds of machines over the years, I can't say that I've ever seen a failure due to the overclock frequency - as long as you are careful to select a frequency at which things are fully stable. I've never had a CPU failure but I've seen plenty of capacitor failures which happen anyway, irrespective of overclocking. I would have repaired 20 or more motherboards (all successfully) which failed due to SCS (swollen capacitor syndrome). Some of those weren't ever overclocked and some were in non-working machines that came my way with the problem obvious when I first tried to fire them up.

The biggest problem is the gradual reduction in cooling efficiency due to accumulating dust and fan bearings that start to 'dry out' with age. Usually the first decent hot spell of spring/summer causes a number of machines to start playing up and if I'm a bit lazy I just reduce the overclock a bit until I get around to doing a bunch of HSF (heat sink/fan) refurbishments. I do a lot of PSU fan jobs (when the PSU starts feeling a little hotter than normal and a fan re-lube with sewing machine oil lasts anywhere from 6 - 18 months usually. Most fans have a removable rubber bung and even those that are completely sealed can be refurbished by drilling a tiny hole and inserting the oil with a syringe.

I'm actually still running the very first machine I started crunching with (Seti classic back in 1999). It's a Celeron 466 overclocked to about 550 or thereabouts. It was my work machine from 1999 to about 2004 and it has run 24/7 ever since I first bought it. It was retired from work duties about the time Seti BOINC was launched and it has continued to crunch to this day (I didn't have the heart to ever turn it off) :-). Everything is still all original - even the hard disk has never missed a beat. It's running Win 98SE and it only ever stops if there is a power failure. The last one was probably about a year ago. Before someone points out that it's impossible to get even two week uptimes (let alone year long ones) with Win98, yes, I know, but I solved that behaviour years ago with the unofficial win98SE service pack (google it if you really must) which completely solved all those stability issues. On Seti it has a RAC of around 40 and around 70K total credit. Before that it did lots of Seti classic WUs. If overclocking per se is so damaging, why is that machine still crunching? :-).

Quote:
Thanks again


You're welcome!

Cheers,
Gary.

MAGIC Quantum Mechanic
MAGIC Quantum M...
Joined: 18 Jan 05
Posts: 1886
Credit: 1403654658
RAC: 1096625

I have been doing this since

I have been doing this since 1999 myself with many different machines (not as many of some of you guys) and I remember on my first machine of the past a PIII 500 how slow and warm the would get in the summer (and a AMD 1600 as I recall) once they died for different reasons I just added them to my pc junk collection since I want more and better although I still run my old P4 2.5 on occasion.

It can get hot in the summer here and mine are all upstairs (I'm sure you all know what that means)

So these days I still run them 24/7 365 but I have the side panel off the boxes so along with the fan on the processors I also just set a small cheap 5in fan next to them blowing on the HD and processors which keeps them plenty cool not to mention I can just look and see if any dust has found it's way inside (I must keep my room clean because I have only had to vacuum out the boxes once or not at all on the newest ones.

But then I didn't bother overclocking the AMD's

One fun thing is just having several different types of processors and comparing them to each other as far as how fast they can send them back and build up those units on the "Pending credit" list

Jean-Marie
Jean-Marie
Joined: 2 Mar 05
Posts: 4
Credit: 1820892
RAC: 0

I've seen at least two

I've seen at least two computers, which seem to abort the GC tasks and which are only executing the ABP tasks. I'm not taking the credits too serious, but still I find that behaviour a bit unfair to the other users.

Regards,
Jean-Marie

metalius
metalius
Joined: 29 Dec 05
Posts: 44
Credit: 173173853
RAC: 0

RE: I've seen at least two

Message 98157 in response to message 98156

Quote:
I've seen at least two computers, which seem to abort the GC tasks and which are only executing the ABP tasks. I'm not taking the credits too serious, but still I find that behaviour a bit unfair to the other users.


Bad boys are not (and will never be) fair players!!!
For example, I know too one user (from Germany), which is aborting all Einstein's main research tasks and running only ABP. No matter, it is written in his profile - he is interested very much in GW research, LOL.
Today I found a bad boy, who found a way, how to write app-info.xml for Einstein@home. So his computer(s) is (are) receiving only ABP.
IMHO, the sentence "Einstein@home is under attack!" is at least partially correct.

Alex
Alex
Joined: 1 Mar 05
Posts: 451
Credit: 507021598
RAC: 115600

RE: RE: I've seen at

Message 98158 in response to message 98157

Quote:
Quote:
I've seen at least two computers, which seem to abort the GC tasks and which are only executing the ABP tasks. I'm not taking the credits too serious, but still I find that behaviour a bit unfair to the other users.

Bad boys are not (and will never be) fair players!!!
For example, I know too one user (from Germany), which is aborting all Einstein's main research tasks and running only ABP. No matter, it is written in his profile - he is interested very much in GW research, LOL.
Today I found a bad boy, who found a way, how to write app-info.xml for Einstein@home. So his computer(s) is (are) receiving only ABP.
IMHO, the sentence "Einstein@home is under attack!" is at least partially correct.

Hi,
if somebody has an app_info.xml telling boinc to download only one specific app I cannot find anything 'bad' in it.
Bad is to delete downloaded apps all the time. This requires extra actions from the servers and delays completing of the work. If the servers and the network have enough 'headroom', nothing else would happen. A problem arises, when more and more users act like that. Fairness is to write a app_info, inhibiting needless network traffic.
But Einstein-Stuff could do something. Some projects offer the choice to check or uncheck, which apps should be loaded.

Regards,
Alexander

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.