Project: 5101 (Run 0, Clone 16, Gen 28) poor cpu loading

[ThunderRd]

I have received one of these WUs and it seems to load the cpu very poorly-only about 65-85 per cent on an Opteron 185 running Ubuntu 8.04.

It also seems to go through stages of both increased and decreased usage; for example, an period of 2 checkpoints per 1%, followed by another period of 3 or 4 checkpoints per 1%. Note the log snippet:

Code: Select all

[09:38:21] + Results successfully sent
[09:38:21] Thank you for your contribution to Folding@Home.
[09:38:21] + Number of Units Completed: 24

[09:38:22] + Sent 1 of 1 completed units to the server
[09:38:22] - Autosend completed
[09:45:55] Timered checkpoint triggered.
[10:00:55] Timered checkpoint triggered.
[10:01:35] Writing local files
[10:01:36] Completed 220000 out of 2000000 steps  (11 percent)
[10:16:35] Timered checkpoint triggered.
[10:31:35] Timered checkpoint triggered.
[10:46:35] Timered checkpoint triggered.
[10:50:33] Writing local files
[10:50:33] Completed 240000 out of 2000000 steps  (12 percent)
[11:05:32] Timered checkpoint triggered.
[11:20:32] Timered checkpoint triggered.
[11:35:32] Timered checkpoint triggered.
[11:44:31] Writing local files
[11:44:31] Completed 260000 out of 2000000 steps  (13 percent)
[11:59:31] Timered checkpoint triggered.
[12:14:31] Timered checkpoint triggered.
[12:29:31] Timered checkpoint triggered.
[12:38:27] Writing local files
[12:38:27] Completed 280000 out of 2000000 steps  (14 percent)
[12:53:27] Timered checkpoint triggered.
[13:08:27] Timered checkpoint triggered.
[13:23:27] Timered checkpoint triggered.
[13:25:49] Writing local files
[13:25:50] Completed 300000 out of 2000000 steps  (15 percent)
[13:40:49] Timered checkpoint triggered.
[13:55:49] Timered checkpoint triggered.
[14:10:11] Writing local files
[14:10:11] Completed 320000 out of 2000000 steps  (16 percent)
[14:25:11] Timered checkpoint triggered.
[14:40:11] Timered checkpoint triggered.
[14:54:14] Writing local files
[14:54:14] Completed 340000 out of 2000000 steps  (17 percent)
[15:09:13] Timered checkpoint triggered.
[15:24:13] Timered checkpoint triggered.
[15:38:22] - Autosending finished units...
[15:38:22] Trying to send all finished work units
[15:38:22] + No unsent completed units remaining.
[15:38:22] - Autosend completed
[15:38:55] Writing local files
[15:38:55] Completed 360000 out of 2000000 steps  (18 percent)
[15:53:54] Timered checkpoint triggered.
[16:08:54] Timered checkpoint triggered.
[16:23:33] Writing local files
[16:23:33] Completed 380000 out of 2000000 steps  (19 percent)
[16:38:32] Timered checkpoint triggered.
[16:53:32] Timered checkpoint triggered.
[17:07:28] Writing local files
[17:07:28] Completed 400000 out of 2000000 steps  (20 percent)
[17:22:28] Timered checkpoint triggered.
[17:37:28] Timered checkpoint triggered.

It doesn't seem to be a problem to run, but is anyone else noticing this on 5101? It's the first one I've had and I'm not sure if it's normal. The box has produced hundreds of WUs previously, so I don't see any reason to think it's the hardware. All of the WUs I have gotten recently have been loading the cpu quite well-normally 92-95%++ at all times.

[ThunderRd]

This unit has mercifully EUE'd at 65%. In the throes of death it was taking 1:19 to finish 1%, something that I have never seen as long as I have been folding, making 5 checkpoints per 1%. I have to believe that something was wrong with it, because I then downloaded a new WU and am right back to normal.

I'd be curious to see if the next donor that receives it has problems; can anyone check if it has been reassigned, out of curiosity?

[Karamiekos]

This unit has mercifully EUE'd at 65%. In the throes of death it was taking 1:19 to finish 1%, something that I have never seen as long as I have been folding, making 5 checkpoints per 1%. I have to believe that something was wrong with it, because I then downloaded a new WU and am right back to normal.

I'd be curious to see if the next donor that receives it has problems; can anyone check if it has been reassigned, out of curiosity?

Never had a 5101, but I have had a 5102.
http://foldingforum.org/viewtopic.php?f ... 102#p50950

Seems to be just thoses projects those in particular from the 51xx based on the old A1 core. Its normal. The EUE isn't, though I have had a problem with the 5102 hanging at 100%....lost a couple WU's that I worked on for 5 days like that, but if you be careful and look up the sticky about saving them, you'll be right as rain.

Wonder why you got EUE....hmmm...

[ThunderRd]

I wouldn't have any question if the output was steady and regular at 1 frame an hour from beginning to end. It's the irregularity that made me wonder; from 30 minutes a frame at some points to 1:19 at some points is quite a large swing. With more than 30 machines running SMP I can't really follow the production of each one very carefully but this one happens to be one of my home machines, and it has done hundreds of SMP WUs without a hiccup or EUE. In fact, it is probably the most reliable box in my farm.

No big problem; I just thought it would be interesting to know how the next guy does with it...

[toTOW]

Someone else completed that WU successfully ...