WU question

Riddick

Does a Early Unit End like this mean im over-oced? im starting to get a little worried as i've got this 2times in a row.. :

[08:46:19] *

---

*
[08:46:19] Folding@home Gromacs Core
[08:46:19] Version 1.48 (May 7, 2003)
[08:46:19]
[08:46:19] Preparing to commence simulation
[08:46:19] - Assembly optimizations manually forced on.
[08:46:19] - Not checking prior termination.
[08:46:21] - Expanded 584481 -> 3018297 (decompressed 516.4 percent)
[08:46:21]
[08:46:21] Project: 368 (Run 8, Clone 28, Gen 16)
[08:46:21]
[08:46:22] Assembly optimizations on if available.
[08:46:22] Entering M.D.
[08:46:44] (Starting from checkpoint)
[08:46:44] Protein: p368_tip5p_pf1_Na
[08:46:44]
[08:46:44] Writing local files
[08:47:21] Completed 78000 out of 100000 steps (78)
[08:47:21] Extra SSE boost OK.
[08:53:26] Gromacs cannot continue further.
[08:53:26] Going to send back what have done.
[08:53:26] logfile size: 46320
[08:53:26] - Writing 46856 bytes of core data to disk...
[08:53:26] ... Done.
[08:53:26]
[08:53:26] Folding@home Core Shutdown: EARLY_UNIT_END
[08:53:29] CoreStatus = 72 (114)
[08:53:29] Sending work to server

seversphere

bring it down a 100 mhz (to be sure) and see if it will work a few frames. It is most likely unstable from the oc. But you could just need to up the vcore or cool the chip better. If you don't want to do either of those then just drop it 50mhz and that should be good.

edit: you should have gromac core version 1.49 by now

Riddick

my client (console) aint downloading 1.49.. in fact another box of mine doesnt either.

i thought that if it was an OC problem it would look like this and not the above :

[06:09:21] *

---

*
[06:09:21] Folding@home Gromacs Core
[06:09:21] Version 1.48 (May 7, 2003)
[06:09:21]
[06:09:21] Preparing to commence simulation
[06:09:21] - Assembly optimizations manually forced on.
[06:09:21] - Not checking prior termination.
[06:09:22] - Expanded 464910 -> 2420513 (decompressed 520.6 percent)
[06:09:22]
[06:09:22] Project: 363 (Run 6, Clone 47, Gen 3)
[06:09:22]
[06:09:22] Assembly optimizations on if available.
[06:09:22] Entering M.D.
[06:09:42] (Starting from checkpoint)
[06:09:42] Protein: p363_tip4p_pf2_Mg
[06:09:42]
[06:09:42] Writing local files
[06:10:19] Completed 35000 out of 250000 steps (14)
[06:10:19] Extra SSE boost OK.
[06:22:34] Writing local files
[06:22:34] Completed 37500 out of 250000 steps (15)
[06:34:49] Writing local files
[06:34:49] Completed 40000 out of 250000 steps (16)
[06:47:03] Writing local files
[06:47:03] Completed 42500 out of 250000 steps (17)
[06:59:19] Writing local files
[06:59:19] Completed 45000 out of 250000 steps (18)
[07:09:01] Quit 101 - Fatal error:
[07:09:01] Step 46959, time 93.918 (ps) LINCS WARNING
[07:09:01] relative constraint deviation after LINCS:
[07:09:01] max 0.000000 (between atoms 1 and 2) rms 1.#QNAN0
[07:09:01]
[07:09:01] Simulation instability has been encountered. The run has entered a
[07:09:01] state from which no further progress can be made.
[07:09:01] If you often see other project units terminating early like this
[07:09:01] too, you may wish to check the stability of your computer (issues
[07:09:01] such as high temperature, overclocking, etc.).
[07:09:01] Going to send back what have done.
[07:09:01] logfile size: 48372
[07:09:01] - Writing 49048 bytes of core data to disk...
[07:09:01] ... Done.
[07:09:01]
[07:09:01] Folding@home Core Shutdown: EARLY_UNIT_END
[07:09:04] CoreStatus = 72 (114)
[07:09:04] Sending work to server

seversphere

over oced can cause any number of problems - some problems might be specific to just that run of wu or client version / core version. Not a big worry if you are running 1.48 - just that some of us got updated. Dropping your speed 100 mhz or even back to default is to check and see if it's Folding or your system that was causing the error. It's most likely because of too high a cpu speed since I assume you are trying to find your max oc right now.

edit: assumed wrong of course :banghead: Are you trying ocing right now or this just started out of the blue? - still drop speed to check and see if it will get thru a wu.

mmonnin

It could be just the WU. You should get partial credit for that last I heard.

Riddick

nah this came out of the blue, i fixed all my file_io_errors with a format to XP . Normally once it receives the WUs this doesnt happen ( same clockspeed as now)

shwaip

If you close F@H and delete teh FAHCore_78.exe file in your folding directory, you'll download the new core. Also, what version of the console are you using?

WuGgaRoO

if u up the vcore a step or two...it should get stable...if u cant do that then i dunno....my friend has instabiity problems for sum unknown reason with his 1800+ regardless of the vcore..it fuxors up

Riddick

shwaip said
. Also, what version of the console are you using?

console 3.24

t1rhino

You could try downloading 3.25 from the beta section.

Straight_Man

If none of the above works, you might want to test your RAM. Bad RAM has done so many weird things that it is not funny. It is a case of NGIBGO (no garbage in,but garbage out) if RAM is not good, is overheated, or if the refresh needs of it at its settings are not being met. Bad RAM can cause values that were calculated right, and not stored in error, to come out full of errors if RAM has major problems or with just a few errors here and there if bad places are tiny or refresh rates are out of sync with rest of system by just a tib..