Sempron not handling folding?

M

Mayhem33

[H]ard|Gawd
Joined
Oct 12, 2004
Messages
1,204
I got a 2800+ sempron that will not run the gromac core for more than a few moments without getting an early end unit. I ran prime for 24 hrs and memtest passed 12 tests with no errors.
It wont run a gromac to save its life though, any ideas?
I havent seen the temp get over 51 degrees celsius. Like I said though it bearly gets started.... corrupts WU.... gets another....and so on. Any ideas?

 
I would guess you would have said this but......has it gotten any tinkers?
 
Mayhem33 said:
I got a 2800+ sempron that will not run the gromac core for more than a few moments without getting an early end unit. I ran prime for 24 hrs and memtest passed 12 tests with no errors.
It wont run a gromac to save its life though, any ideas?
I havent seen the temp get over 51 degrees celsius. Like I said though it bearly gets started.... corrupts WU.... gets another....and so on. Any ideas?

Click to expand...

any particular protein or does it throw up after a couple of minutes?
 
Spectre said:
I would guess you would have said this but......has it gotten any tinkers?
Click to expand...
It has tinkers now that i set it for timeless.....and it runs those fine so far. :rolleyes:

magnusvir said:
any particular protein or does it throw up after a couple of minutes?
Click to expand...
The way I had it configured it was only getting those 600 point gro's...but yes it choked on them right away. :confused:

Any ideas?

 
Mayhem33 said:
It has tinkers now that i set it for timeless.....and it runs those fine so far. :rolleyes:
Click to expand...

I was going to suggest trying that if you hadn't already.....good call.


The way I had it configured it was only getting those 600 point gro's...but yes it choked on them right away. :confused:

Any ideas?

Click to expand...

From what I understand some of those 600 pointers are beta units.....so they may not be stable on all configs....but I could be wrong. The 2.8ghz P4 I have set for big WU's loves 600 pointers.........so could be something to do with differences in architecture.
 
Spectre said:
I was going to suggest trying that if you hadn't already.....good call.




From what I understand some of those 600 pointers are beta units.....so they may not be stable on all configs....but I could be wrong. The 2.8ghz P4 I have set for big WU's loves 600 pointers.........so could be something to do with differences in architecture.
Click to expand...

True some of those are beta...but the number I was throwing away....not possible! :(
I have another sempron chip chewing those 600 pointers all day long on a stock HSF....this one will not work for me. The difference between the 2 sempron chips the good is a 1.5 at 1.8
and the bad....is stock at 2.0. I even have a 60 mm 5800 RPM fan on the bad one and still....
no luck!

 
Mayhem33 said:
True some of those are beta...but the number I was throwing away....not possible! :(
I have another sempron chip chewing those 600 pointers all day long on a stock HSF....this one will not work for me. The difference between the 2 sempron chips the good is a 1.5 at 1.8
and the bad....is stock at 2.0. I even have a 60 mm 5800 RPM fan on the bad one and still....
no luck!

Click to expand...

Well in that case I am stumped. Sorry maybe some one else will see this and be able to help more. Good luck though.
 
Mayhem33 said:
True some of those are beta...but the number I was throwing away....not possible! :(
I have another sempron chip chewing those 600 pointers all day long on a stock HSF....this one will not work for me. The difference between the 2 sempron chips the good is a 1.5 at 1.8
and the bad....is stock at 2.0. I even have a 60 mm 5800 RPM fan on the bad one and still....
no luck!

Click to expand...

If it helps you any, I have had 5 of those beta p1475 units on my Athlon 64. Every one of them threw up at different points in time. 1 threw up after 2 steps, another after 45 steps. I have not had a problem yet with the 600 pointers, but maybe you are getting the bad runs.
 
magnusvir said:
If it helps you any, I have had 5 of those beta p1475 units on my Athlon 64. Every one of them threw up at different points in time. 1 threw up after 2 steps, another after 45 steps. I have not had a problem yet with the 600 pointers, but maybe you are getting the bad runs.
Click to expand...

ROTFL!
mdr.gif

I got the bad runs!!....look out!
flush.gif


 
I did just try updating the bios....but im scared! I dont wanna trash any more WU's! :eek:
What to do?
nuke.gif


 
I have a Sempron 2200+ clocked to 1.67GHz.... it doesn't have a problem with WU, but for some reason it refuses to pick up a 600 pointer.... I have it set to big packets, and it still gets the traditional 30-90 point gromacs.... it has grabbed a few 241point Tinkers though... did well on those....


Keep on Folding!!

 
I have a 1 Ghz Morgan core Duron currently chewing away on a 600 point Gromac and it's had no trouble so far. It should be done with it sometime this morning or early afternoon. I don't remember what it's working on and since I'm at work I can't check.

The CPU in my sig has had trouble with a couple of 600 pointers also although it's been a while. I think I have probably just gotten lucky with some of the WUs I have received. After a couple of early ones I got, I have had no trouble really.

To recap, I don't think it's an architecture problem. You may have gotten a bum CPU. If you have anything else that utilizes SSE heavily, you may want to try that and see if it craps out also. In that case, I would think that there is something wrong with the SSE part of the processor.

 
OSUguy98 said:
I have a Sempron 2200+ clocked to 1.67GHz.... it doesn't have a problem with WU, but for some reason it refuses to pick up a 600 pointer.... I have it set to big packets, and it still gets the traditional 30-90 point gromacs.... it has grabbed a few 241point Tinkers though... did well on those....


Keep on Folding!!

Click to expand...

You have to have enough Megs O Ram reporting to Stanford to be assigned to the big units server. The client reports the amount of memory before each new WU assignment.

Code: 
[18:55:22] + Attempting to get work packet
[18:55:22] - Will indicate memory of 255 MB.
[18:55:22] - Connecting to assignment server
[18:55:22] Connecting to [url]http://assign.stanford.edu:8080/[/url]
[18:55:23] Posted data.
[18:55:23] Initial: 41AB; - Successful: assigned to (171.65.103.156).

If you have 256 megs in a system with onboard video and have the shared memory set too high you will fail to get assigned to .156 because your reported memory will be too low. If you only have 128 megs, it will never get a .156 assignment regardless of onboard video with shared memory or not.

 
Mayhem33 said:
I did just try updating the bios....but im scared! I dont wanna trash any more WU's! :eek:
What to do?
nuke.gif


Click to expand...

Test the memory or try a different stick. The problem may not be CPU or MB related at all. Since the big units use more memory than anything we have ever had before, RAM problems that may have never shown up may try to put a damper on the folding party.

 
Killer[MoB] said:
Test the memory or try a different stick. The problem may not be CPU or MB related at all. Since the big units use more memory than anything we have ever had before, RAM problems that may have never shown up may try to put a damper on the folding party.

Click to expand...

Mayhem33 said:
I got a 2800+ sempron that will not run the gromac core for more than a few moments without getting an early end unit. I ran prime for 24 hrs and memtest passed 12 tests with no errors.
Click to expand...
I did twelve complete runs with memtest...no probs?

 
Varroa said:
I have a 2800+ Sempron and it works with those no problem!
Click to expand...
Thats what I mean...it is strange! I think someone was on the right track when they said I got a bum cpu! :(

 
Ahh take the whole computer apart and clean it up. By time you put it back together, you'll have broken something, need a new something then have a whole new puter.

Problem solved.
 
marty9876 said:
Ahh take the whole computer apart and clean it up. By time you put it back together, you'll have broken something, need a new something then have a whole new puter.

Problem solved.
Click to expand...
Were you spying on me this weekend? HA Ha thats sorta how it went. This box is for my mom and she was going to get my old rig.....now hers is all brand new! :eek:
Which is where the sempron chip comes in....[sigh] :rolleyes:
Tuns out the 1 gig T-Bird that was working just up and died!

 
I did click the link from the other thread. It was a joke. We are going to piss off DK_??, he deleted some other posts in that thead for being off topic.

So I figured I'd crap in here.... :)

Ohh, on this puter... umm maybe relax the memory timings..

Seriously, start with the basics and give the boxen a good once over with a fresh "eye" ?
 
marty9876 said:
I did click the link from the other thread. It was a joke. We are going to piss off DK_??, he deleted some other posts in that thead for being off topic.

So I figured I'd crap in here.... :)

Ohh, on this puter... umm maybe relax the memory timings..

Seriously, start with the basics and give the boxen a good once over with a fresh "eye" ?
Click to expand...
I tried leaving most everything stock to see if I was screwing things up somehow but it didnt make any difference. The mem is Corsair value and set to SPD.
I been wanting to give it a once over all right!!!!!!!!! First a good kick and then the hammer!!!
Yes get the hammer!

 
Mayhem33 said:
I tried leaving most everything stock to see if I was screwing things up somehow but it didnt make any difference. The mem is Corsair value and set to SPD.
I been wanting to give it a once over all right!!!!!!!!! First a good kick and then the hammer!!!
Yes get the hammer!

Click to expand...

I had a box where SPD was too high for the memory, even thou it matched the stupid sticks... Turned out to me a "weak" mother board.
 
marty9876 said:
I had a box where SPD was too high for the memory, even thou it matched the stupid sticks... Turned out to me a "weak" mother board.
Click to expand...
I thought the same thing too and got another mobo but still having the same trouble.
Seems to not like the SSE instructions that come on those gromac WU.
Runs those tinkers just fine! It does seem a waste to me though to have that much horsepower set to timeless!

 
Go to the EM-DC.com web site and go to the download section. At or near the bottom is an SSE test program which will thrash the CPU ... if it fails your CPU probably has an issue from everything else posted here... Those memory test programs don't hammer the CPU like this test program or a real F@H Gromacs core will.
 
LPerry said:
Go to the EM-DC.com web site and go to the download section. At or near the bottom is an SSE test program which will thrash the CPU ... if it fails your CPU probably has an issue from everything else posted here... Those memory test programs don't hammer the CPU like this test program or a real F@H Gromacs core will.
Click to expand...

Cool program LPerry, I was looking for something like that on the net...and no luck. That really helps and glade to know you have that!
I ran the program for 30-40 minutes and no problem, it got to a temp of 52 celcius on the proccessor but it should be ok there. So much for my SSE theroy!
I did update the bios on the mobo and re-configed F@H so I would get the gro's again. It is working a tinker at the moment, so I will post back with an update.
Thanks again to all! :cool:

 
marty9876 said:
Ohh, on this puter... umm maybe relax the memory timings..

Seriously, start with the basics and give the boxen a good once over with a fresh "eye" ?
Click to expand...
Hey Marty, I gave my idea about the SSE a test and it passed no probs. So I did as you sugested and started looking through the box agian and it turns out the mem is Corsair XMS 2400 and I think that comes out to a bus of 150 or 300 DDR. The CPU needs to run at 266 to run at its correct speed of 2 gigs, so I am trying to ease up on the timings and see if that will work.
I think the timmings are 2/3/3/6 and going to try 2.5/3/3/7. Whats your guess on what they should be?

 
Going to single channel memory seemed to solve my problem....which was similar.


 
Hito Bahadur said:
Going to single channel memory seemed to solve my problem....which was similar.


Click to expand...

You are talking about going from slots 1 and 3 to like 1 and 2 right? So you are talking about populating the dimms so they don't run in a dual array.
This solved your problems?
What board are you talking about?

 
Mayhem33 said:
You are talking about going from slots 1 and 3 to like 1 and 2 right? So you are talking about populating the dimms so they don't run in a dual array.
This solved your problems?
What board are you talking about?

Click to expand...

In my case, a Shuttle XPS, but if its an Nforce2 board, it can run in Single Channel... i.e. 1 stick of memory, which is what I did. I went from 2x512 to 1x512.
 
Hito Bahadur said:
In my case, a Shuttle XPS, but if its an Nforce2 board, it can run in Single Channel... i.e. 1 stick of memory, which is what I did. I went from 2x512 to 1x512.
Click to expand...
Well if timmings dont fix it I will give that a try!
Yes I have a Nforce board as well! Abit NF7-S2G you can also get those to run in single channel by ussing slot 1 and 2. Those run on the same channel slot 3 is another channel though. I am useing 2 256 meg sticks. Gotta keep them both in there, XP likes its mem!
Thanks though!

 
Mayhem33 said:
Hey Marty, I gave my idea about the SSE a test and it passed no probs. So I did as you sugested and started looking through the box agian and it turns out the mem is Corsair XMS 2400 and I think that comes out to a bus of 150 or 300 DDR. The CPU needs to run at 266 to run at its correct speed of 2 gigs, so I am trying to ease up on the timings and see if that will work.
I think the timmings are 2/3/3/6 and going to try 2.5/3/3/7. Whats your guess on what they should be?

Click to expand...

I believe all AthlonXP-based Semprons run 166/333Mhz bus, so your 2400 rated memory is too low. However, since it passed memtest, it's probably not the culprit. Try running in single-channel like others suggested, though.
 
Mayhem33 said:
I did twelve complete runs with memtest...no probs?

Click to expand...

I usually try to let memtest run for 24hrs. I have had many problems not come until 11+ hrs of stress testing (this allows for a full days fluctuation in external temps, sunlight, etc. ). Same with prime 95.

Man, it seems like the longer that I am in school and the more damn scientific journal articles I read the worse my writing is getting.
 
Hito Bahadur said:
Going to single channel memory seemed to solve my problem....which was similar.


Click to expand...

You were right on the money Hito Bahadur, moved mem into single channel mode and all is well!!!
Thanks for the info!

 
Mayhem33 said:
You were right on the money Hito Bahadur, moved mem into single channel mode and all is well!!!
Thanks for the info!

Click to expand...


No problem. It's good to hear. Thanks for letting me know.


 
You must log in or register to reply here.
Back
Top