• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

F@H - Giant Gromacs units roaming the wilderness

H

Hyperfocal

Senior member
Just a heads up.

There are some huge new Gromacs units out there.

P857, p858 and p859 are all 160 point units that take a LONG, LONG time to crunch.

It's looking like nearly 40 hours on my 3200+ for p858_p53dimer858. :Q
 
Aww carp... I'm crunching the same one on my 2500@3200. Oh well more points 😛.

Current Work Unit
-----------------
Name: p858_p53dimer858
Download time: March 19 21:55:06
Due time: July 17 21:55:06
Progress: 4% [__________]

EDIT: My laptop seems to be chewing right through the same unit. It's at 54%, downloaded at about 3AM today according to the unitinfo file...
 
I dunno which Gromac unit my 2500+ running at 2.3GHz was working on earlier, but it was clicking by each step in about 3 mins.
 
they were 160 points, they have been changed to 120 points...this when I have 3 rigs working on them.

Of course I got them on my 3 slowest rigs, frame times 28 min, 29 min, and 35 min.

I just updated the client list in FAH StatsLog, and these dropped to 120 points.....:brokenheart:
 
These were Beta only WUs on the 18th, they are now getting to public computers.

Stanford has changed the points as well as the deadline to return them a time or 2 already, there may be more changes in store for these as they are very fresh.

lots of then out there , I have 9 rigs on them now.......
 
I was just notified on the Stanford beta fourm that they are in fact still 160 points, just the deadline was changed.
Seems to be a bug with the FAH LogStats program, prob be corrected in a day or so.
 
must be all they are sending out to the servers now, just got another one...10 out of 16 rigs now on these monsters.
 
[20:33:08] *------------------------------*
[20:33:08] Folding@home Gromacs Core
[20:33:08] Version 1.56 (February 2, 2004)
[20:33:08]
[20:33:08] Preparing to commence simulation
[20:33:08] - Assembly optimizations manually forced on.
[20:33:08] - Not checking prior termination.
[20:33:10] - Expanded 328566 -> 1861229 (decompressed 566.4 percent)
[20:33:10] - Starting from initial work packet
[20:33:10]
[20:33:10] Project: 858 (Run 109, Clone 10, Gen 0)
[20:33:10]
[20:33:10] Assembly optimizations on if available.
[20:33:10] Entering M.D.
[20:33:16] Protein: p858_p53dimer858
[20:33:16]
[20:33:17] Writing local files
[20:33:17] Extra SSE boost OK.
[20:33:17] Writing local files
[20:33:19] Completed 0 out of 2500000 steps (0)
[21:27:53] Writing local files
[21:27:55] Completed 25000 out of 2500000 steps (1)
[22:16:53] Writing local files
[22:16:55] Completed 50000 out of 2500000 steps (2)
[23:00:49] Writing local files
[23:00:51] Completed 75000 out of 2500000 steps (3
Click to expand...
woah......almost an hour to crunch them on my 1700+
 
These WUs must like big L2 caches, my laptop (A64 @ 2.0 with 1MB L2) is chugging right through them, but my desktop (XP @ 2.2) isn't faring quite so well.
 
Originally posted by: MDE
These WUs must like big L2 caches, my laptop (A64 @ 2.0 with 1MB L2) is chugging right through them, but my desktop (XP @ 2.2) isn't faring quite so well.
Click to expand...

What are your times on the A64? My prescott is running them in 19 min.

[05:15:58] Project: 858 (Run 223, Clone 3, Gen 0)
[05:15:58]
[05:15:58] Assembly optimizations on if available.
[05:15:58] Entering M.D.
[05:16:19] (Starting from checkpoint)
[05:16:19] Protein: p858_p53dimer858
[05:16:19]
[05:16:19] Writing local files
[05:16:19] Completed 75000 out of 2500000 steps (3)
[05:16:19] Extra SSE boost OK.
[05:35:08] Writing local files
[05:35:10] Completed 100000 out of 2500000 steps (4)
[05:54:00] Writing local files
[05:54:02] Completed 125000 out of 2500000 steps (5)
 
Here is my daughters 2.4C @ 3gig

04:16:04] Protein: p858_p53dimer858
[04:16:04]
[04:16:04] Writing local files
[04:16:04] Completed 1712006 out of 2500000 steps (68)
[04:16:05] Extra SSE boost OK.
[04:28:00] Writing local files
[04:28:02] Completed 1725000 out of 2500000 steps (69)
[04:50:57] Writing local files
[04:50:59] Completed 1750000 out of 2500000 steps (70)
 
About 22 minutes

[03:23:28] Completed 1500000 out of 2500000 steps (60)
[03:45:14] Writing local files
[03:45:17] Completed 1525000 out of 2500000 steps (61)
[04:06:57] Writing local files
[04:06:59] Completed 1550000 out of 2500000 steps (62)
[04:28:37] Writing local files
[04:28:39] Completed 1575000 out of 2500000 steps (63)
[04:50:16] Writing local files
[04:50:18] Completed 1600000 out of 2500000 steps (64)
[05:12:27] Writing local files
[05:12:29] Completed 1625000 out of 2500000 steps (65)
 
3200+

Protein: p858_p53dimer858

[03:42:48] Writing local files
[03:42:50] Completed 1325000 out of 2500000 steps (53)
[04:05:45] Writing local files
[04:05:47] Completed 1350000 out of 2500000 steps (54)
[04:34:46] Writing local files
[04:34:49] Completed 1375000 out of 2500000 steps (55)
[04:57:47] Writing local files
[04:57:49] Completed 1400000 out of 2500000 steps (56)

Duron @ 2ghz

Protein: p859_p53dimer859

[04:23:23] Writing local files
[04:23:25] Completed 1275000 out of 2500000 steps (51)
[04:52:40] Writing local files
[04:52:42] Completed 1300000 out of 2500000 steps (52)

I can't wait to get one on my celeron 900
rolleye.gif
 
I've got an 858 turning in 4 min + 20 sec or so...not too bad, I say.

Just a tip: are you guys sure your amd's have that wonderful -forceSSE option on? it gives you around a 30-50% speed increase over the default 3dnow version...
 
Originally posted by: osage
a p858_p53dimer858 ? ?

what cpu are you getting 4 min + 20 sec frame times with ?
Click to expand...

Thats what I thought! Is there a possibility that its 1 HOUR and 4 minutes 20 seconds?? 😉
 
Originally posted by: osage
a p858_p53dimer858 ? ?

what cpu are you getting 4 min + 20 sec frame times with ?
Click to expand...

Maybe one of those Michael's superfast computers😛

I just got a p858 on my old celery 1hour 31 minutes per frame. Nearly 6 1/2 days.:Q

4:20 sound more like a p848 or p853. I have -forceSSE on all my rigs.
 
Originally posted by: Hyperfocal
Originally posted by: osage
a p858_p53dimer858 ? ?

what cpu are you getting 4 min + 20 sec frame times with ?
Click to expand...

Maybe one of those Michael's superfast computers😛

I just got a p858 on my old celery 1hour 31 minutes per frame. Nearly 6 1/2 days.:Q

4:20 sound more like a p848 or p853. I have -forceSSE on all my rigs.
Click to expand...

It's taking 1hr 26min on my work pc a P3 800.
 
ahhh, so im not the only one to encounter this monster.

using the info about how they come up with dealines...it looks like it took their celery 500 around 16 days to kill this one off.
ouch.
 
I guess I've got an apology...I must have made a mistake in my calculations
here is a snippet from my log:

[12:44:18] Protein: p858_p53dimer858
[12:44:18]
[12:44:18] Writing local files
[12:44:18] Completed 675000 out of 2500000 steps (27%)
[12:44:19] Extra SSE boost OK.
[13:18:55] Writing local files
[13:18:57] Completed 700000 out of 2500000 steps (28%)
[13:53:26] Writing local files
[13:53:28] Completed 725000 out of 2500000 steps (29%)
 
Aww crap. I got a few of them now. Here's my frame times.

PIII 1Ghz - 1 hour 4 min - 4 days 10 hours left
P4 2.6Ghz - 21 min 15 sec - 1 day 10 hours left
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top