???what the ???

[cadaveca]

[03:23:19] Completed 760000 out of 1000000 steps (76)
[03:31:47] + Paused
[03:31:47] Writing local files
[03:31:47] Suspending work thread...

Folding@Home Client Shutdown.


--- Opening Log file [September 13 12:18:23]


# Windows Graphical Edition ################################################## #
################################################## #############################

Folding@Home Client Version 5.02

http://folding.stanford.edu

################################################## #############################
################################################## #############################

Launch directory: C:\Program Files\Folding@Home
Arguments: -advmethods

[12:18:23] - Ask before connecting: No
[12:18:23] - Use IE connection settings: Yes
[12:18:23] - User name: cadaveca (Team 45)
[12:18:23] - User ID: 3CF432EA643A19DE
[12:18:23] - Machine ID: 1
[12:18:23]
[12:18:23] Loaded queue successfully.
[12:18:23] Initialization complete
[12:18:23] + Benchmarking ...
[12:18:26]
[12:18:26] + Processing work unit
[12:18:26] Core required: FahCore_78.exe
[12:18:26] Core found.
[12:18:26] Working on Unit 03 [September 13 12:18:26]
[12:18:26] + Working ...
[12:18:27]
[12:18:27] *------------------------------*
[12:18:27] Folding@home Gromacs Core
[12:18:27] Version 1.67 (July 13, 2004)
[12:18:27]
[12:18:27] Preparing to commence simulation
[12:18:27] - Ensuring status. Please wait.
[12:18:44] - Looking at optimizations...
[12:18:44] - Working with standard loops on this execution.
[12:18:44] - Previous termination of core was improper.
[12:18:44] - Files status OK
[12:18:45] - Expanded 223821 -> 1072601 (decompressed 479.2 percent)
[12:18:45] - Could not read from checksum file
[12:18:45] - Starting from initial work packet
[12:18:45]
[12:18:45] Project: 214 (Run 1, Clone 27, Gen 3)



i paused the core and shut down as per usual...upon restart of my machine, i found this in my log. so i guess i lost a almost complete unit. a 238 pointer. and the HDD died on the other machine losing the 238 pointer there too...second machine seems to be running fine now, and has since completed the first WU on the new HDD, but this machine seems flaky at folding now at best.. instead of it taking 8 minutes to complete a frame on a sub 100 pointer, it's taking 17...any ideas as to what's going on?

[Gizmo]

What CPU? It isn't using SSE optimizations or anything, but the WU is a Gromacs.

[cadaveca]

the mobile athlon barton

[Gizmo]

Then you definitely have a problem, because it should be detecting and using SSE. That's why your frame times are so high.

[cadaveca]

deleted the core and downloaded new one...we'll see how that goes...

[Gizmo]

It might actually be working without SSE because of the way it terminated the first time. I'm not too sure about that. Maybe Allan has some thoughts?

[Allan]

cadaveca, please email me the entire contents of the logfile and I'll have a closer look at it. Thx!

allan at aoaforums.com

[Toro]

i think i have a problem with one of mine not detecting SSE. cant get the motherboard to set the CPU up properly, and so Windows doesnt either :|


its a work in progress!

[cadaveca]

Originally Posted by Allan

cadaveca, please email me the entire contents of the logfile and I'll have a closer look at it. Thx!

allan at aoaforums.com

rgr...will be in two parts.

[Allan]

No problemo.

[Allan]

From the looks of it, it discarded the WU. This could possibly be because by using the -advmethods flag you are requesting beta-units (well, at least sort of beta) and you will sometimes encounter units of this sort that have errors.

I can see from your log file that it is using SSE again, so I don't think there is anything you need to worry about. You can remove the -advmethods flag if you do not wish to recieve any more of these WUs.

[Staz]

I was getting lots and lots of discarded work units, but it seems to have fixed itself. My production went down to about 1/4th of what it was for about 10 days but seems to be back to normal now.

[cadaveca]

i redownloaded the core...maybe this is what has fixed the issue...
regardless, the next WU is about 1/2 done, so we will see what happens.