BPs in short supply once again?

[Steveo989]

On all of my computers I have only recieved one BP and that was approx a week ago. My Dothan is getting hungry for BPs :p

[pik4chu]

honestly i dont know how toi tell :)

[Bon3thugz43v3r]

You want QMD's for those.

Edit...
Steveo I don't know if you have noticed but you've been a member here since December 2004 and you've made 155 threads. I have been a member since October of 2002 and have made 186 threads. Is this a good thiing? :shrug:

[Silver]

Nothing but little stuff here as well. I mean time consuming low pointage gromacs. To think I thought the tinkers were bad.

[dicecca112]

I'm continuously getting QMDs

[Silver]

As do I on the Intels. AMD on the other hand is getting short changed big time.

[Steveo989]

You want QMD's for those.

Edit...
Steveo I don't know if you have noticed but you've been a member here since December 2004 and you've made 155 threads. I have been a member since October of 2002 and have made 186 threads. Is this a good thiing? :shrug:
I don't know why it matters or why you would pay attention, for all I care you could have 1000 threads and I wouldn't say anything about it. I try to keep the boards active a bit with some ?'s and stuff :temper:

[mjw21a]

Woah, if you're running AMD's then I'm glad that you've no Intel's Silver. You'll be overtaking me very soon now as it is anyways..... ;o)

Ah well, I guess its all for a good cause. Fold on man.

[seadave77]

Edit...
Steveo I don't know if you have noticed but you've been a member here since December 2004 and you've made 155 threads. I have been a member since October of 2002 and have made 186 threads. Is this a good thiing?


I've been around since July '02 and have made 38 threads. What's your point? He asked a question, so why bring up his thread history?

Steve: I don't know about the BP's. I don't pay attention to what I get as long as I'm folding!

[Silver]

Woah, if you're running AMD's then I'm glad that you've no Intel's Silver. You'll be overtaking me very soon now as it is anyways..... ;o)

Ah well, I guess its all for a good cause. Fold on man.

I run both, two intels and two amds. Allows me the luxury of posting in both camps. :p

[MLMIB]

I started about 7 new instances today(mostly just upgrading versions on old borgs) 'n they all got BP's if I had it turned on(used up a heck of a lot of ram, glad they all have a gig

[Bon3thugz43v3r]

Just mentoning that theirs a lot of pointles threads...

[walaka7]

Well the best you can hope for is BP's as you cant get QMD's on the dothan anyway (stanford considers them P3's ATM) The servers that distribute BP's are up, you are just missing the rotation at this time.

I see your point bon3, but perhaps this was a point better made when all the laptop threads were hemoraging the folding forums :D

[ChasR]

You want QMD's for those.

Edit...
Steveo I don't know if you have noticed but you've been a member here since December 2004 and you've made 155 threads. I have been a member since October of 2002 and have made 186 threads. Is this a good thiing? :shrug:

Dothans can't fold QMDs. The QMD core recognizes them as a P3 which doesn't have SSE2.

[mjw21a]

I run both, two intels and two amds. Allows me the luxury of posting in both camps. :p

Hmmmm, I believe that would give you the best of both worlds. No matter what Stanford throws at you it can't hurt you're production too much. A smart move I'd say.


Dothans can't fold QMDs. The QMD core recognizes them as a P3 which doesn't have SSE2.

It'll just be a matter of time before Stanford fixes this and both Dothans and Athlon 64's will be able to fold QMD cores. :shrug:

[Steveo989]

Dothan's at least because they are intels after all.

[stratcatprowlin]

Production will go through the roof when A64's can fold QMD's.I'm not saying they will fold any better than Intel's(I have no idea) but just being able to get access to those points will be fantastic! I can't wait.

[TollhouseFrank]

it'd be nice if QMD's would come to AMD CPU's... i'd almost rather that they made a 450 point WU that only AMD could fold... so that each CPU would be equal in output

[ChasR]

ATM there are the following big WUs available:


16,574 QMDs (p19xx)
172 p147x
71 p13xx
203 600 pointers (p1140, p1141, p1144, p1149, p1150)

Are big packets in short supply? Yes if you've got an AMD rig, No if you've got an Intel P4 w/512 (or more)MB rig.

[Steveo989]

My dothan got it's hands on a 600pointer, time to own some chesseheads :p

[ChasR]

My dothan got it's hands on a 600pointer, time to own some chesseheads :p

You'll need the p147x 364 pointers to own them.

[Steveo989]

I just finished one of those, it took only 12 hours to do it too.

[mjw21a]

See how far that lappy of yours will stably overclock Steveo......

[Steveo989]

How do I overclock it, it's a gateway and I thought major retailors lock BIOS to change FSB

[Silver]

clockgen maybe?

[Steveo989]

what web site?

[mjw21a]

You can find ClockGen here: http://www.cpuid.org/clockgen.php

Happy overclocking.

[Steveo989]

Does not support whatever mainboard I have.

[mjw21a]

That sucks......

[ChasR]

I have doubts it's safe to OC a laptop very far. You'll get into thermal throttling with any significant increase in vcore (if any even allow adjustment in BIOS). I presume our Dothan OCers are using ASUS motherboards with the ASUS Pentium M adapter in desktops. Perhaps one of them can pipe in?

[Mark620]

Big packets actually give me fewer PPD than tinkers on my AMD machines.

[TollhouseFrank]

I've noticed that as well, Mark. Tinkers fly by on my machines... i can turn in 3- 241point tinkers in the time it takes me to fold 1 BP... 3x241 is 723... I earn more PPD on tinkers myself

[ChasR]

Big packets actually give me fewer PPD than tinkers on my AMD machines.

The data I've collected on quite a few AMD rigs, mostly XP-Bartons, but also an A64 and an FX-53 do not support your assertion. I'm not including data on the FX-53 or A64 winnie but trust me the difference between Tinker and Big WUs is far more pronounced on them, in favor of Big WUs, than on the XPs. Here is some of that data:

XP-M Barton

Comp Name Speed Protien Core Value Frames Frame Time Total Time PPD PPW ppd/GHz
Marie 2400 p1273_proj1273_A21ext_d15_99phi Gromac 83 100 0:10:41 17:48:20 111.9 783 46.61
WS19-Matt 2500 p1108_Kiefhaber_20 Tinker 37 40 0:11:20 7:33:29 117.5 822 47.00
Marie 2400 p1278 Gromac 59 100 0:07:22 12:16:30 115.4 807 48.07
ChasR 2532 p1286_A21nat_ntxd15 Gromac 73 100 0:08:22 13:57:08 125.6 879 49.59
ChasR 2532 p1281_A21ext_d13_99phi Gromac 72 100 0:08:12 13:39:40 126.5 885 49.96
Marie 2400 p1286_A21nat_ntxd15 Gromac 73 100 0:08:44 14:32:58 120.4 843 50.17
WS19-Matt 2500 p1286_A21nat_ntxd15 Gromac 73 100 0:08:19 13:51:05 126.5 885 50.59
ChasR 2532 p741_Villin-pfold-protein-confined Gromac 149 100 0:16:39 27:45:17 128.8 902 50.89
ChasR 2532 p234 Gromac 52 100 0:05:48 9:39:28 129.2 905 51.04
Marie 2400 p741_Villin-pfold-protein-confined Gromac 149 100 0:17:12 28:39:20 124.8 874 52.00
WS19-Matt 2500 p741_Villin-pfold-protein-confined Gromac 149 100 0:16:26 27:23:33 130.5 914 52.22
WS19-Matt 2500 p1593_LZIP-1599-1-48_pfold Gromac 57 100 0:06:12 10:19:50 132.4 927 52.97
Marie 2400 p1605_vill8MUre99p Gromac 78 100 0:08:49 14:41:30 127.4 892 53.09
WS19-Matt 2500 p2002_ab1-43_4mer_opls Gromac 180 100 0:18:50 31:23:40 137.6 963 55.04
Philip 2600 p2002_ab1-43_4mer_opls Gromac 180 100 0:17:48 29:39:50 145.6 1019 56.01
Marie 2400 p742_Villin-pfold-protein-confined Gromac 178 100 0:18:50 31:23:00 136.1 953 56.72
Marie 2400 p957_pa_drf2 Gromac 48 100 0:04:49 8:01:10 143.7 1006 59.85
WS19-Matt 2500 p957_pa_drf2 Gromac 48 100 0:04:35 7:39:00 150.6 1054 60.24
Marie 2400 p959 Gromac 48 100 0:04:46 7:57:00 144.9 1014 60.38
Marie 2400 p973_fkcomp_dock_M24 Dgromac 84 100 0:06:35 10:58:00 183.8 1287 76.60
ChasR 2532 p1145_L939_WT_298K Tinker 241 400 0:04:13 28:03:48 206.1 1443 81.40
ChasR 2532 p1146_L939_WT_298K Tinker 241 400 0:04:10 27:46:44 208.2 1458 82.23
ChasR 2532 p1147_L939_WT_298K Tinker 241 400 0:04:10 27:44:44 208.5 1459 82.33
WS19-Matt 2500 p638_L939_K12M_ext Tinker 239 400 0:04:08 27:31:28 208.4 1459 83.36
ChasR 2532 p1137_p1131_L939_K12M_355K Tinker 241 400 0:04:03 27:02:08 213.9 1498 84.49
ChasR 2532 p1136_p1130_L939_K12M_355K Tinker 241 400 0:04:03 26:59:24 214.3 1500 84.64
Marie 2400 p1136_p1130_L939_K12M_355K Tinker 241 400 0:04:14 28:16:04 204.6 1432 85.26
Marie 2400 p1136_p1130_L939_K12M_355K Tinker 241 400 0:04:14 28:12:08 205.1 1436 85.45
Marie 2400 p1136_p1130_L939_K12M_355K Tinker 241 400 0:04:13 28:08:16 205.6 1439 85.65
Marie 2400 p1136_p1130_L939_K12M_355K Tinker 241 400 0:04:12 27:59:20 206.7 1447 86.11
ChasR 2532 p1151_L939_K12M_298K_DT2ns Tinker 241 400 0:03:59 26:30:28 218.2 1527 86.18
Marie 2400 p1136_p1130_L939_K12M_355K Tinker 241 400 0:04:07 27:27:20 210.7 1475 87.78
WS19-Matt 2500 p1139_p1133_L939_K12M_355K Tinker 241 400 0:03:55 26:05:28 221.7 1552 88.67
ChasR 2532 p1138_p1132_L939_K12M_355K Tinker 241 400 0:03:51 25:41:16 225.2 1576 88.93
Philip 2600 p1323_1fnt_a25_9.1ps-1_uf Gromac (Big) 308 100 0:16:47 27:57:39 264.4 1851 101.68
ChasR 2532 p1325_1fnt_a40_9.1ps-1_uf Gromac (Big) 310 100 0:17:11 28:37:40 259.9 1819 102.64
Philip 2600 p590_BBA5_pf Tinker 93 250 0:02:00 8:19:24 268.2 1877 103.14
WS19-Matt 2500 p590 Tinker 93 250 0:02:04 8:37:01 259.0 1813 103.61
WS19-Matt 2500 p590 Tinker 93 250 0:02:04 8:35:50 259.6 1817 103.85
Philip 2600 p1144_RIBO_nopeptide Gromac (Big) 600 100 0:31:50 53:03:40 271.4 1900 104.38
Philip 2600 p1144_RIBO_nopeptide Gromac (Big) 600 100 0:31:46 52:56:57 272.0 1904 104.60
ChasR 2532 p1150_RIBO_semihelixfrom1141 Gromac (Big) 600 100 0:31:31 52:32:17 274.1 1919 108.25
ChasR 2532 p1150_RIBO_semihelixfrom1141 Gromac (Big) 600 100 0:31:22 52:17:20 275.4 1928 108.77
ChasR 2532 p1150_RIBO_semihelixfrom1141 Gromac (Big) 600 100 0:30:44 51:14:00 281.1 1967 111.01
ChasR 2532 p1141_RIBO_FSpeptide_HEL_nospring Gromac (Big) 600 100 0:30:43 51:12:10 281.2 1969 111.07
ChasR 2532 p1140_RIBO_FSpeptide_EXT_nospring Gromac (Big) 600 100 0:30:30 50:50:00 283.3 1983 111.88
Marie 2400 p1141 checkpoint 15 Gromac (Big) 600 100 0:32:10 53:36:40 268.6 1880 111.92
WS19-Matt 2500 p1141_RIBO_FSpeptide_HEL_nospring Gromac (Big) 600 100 0:30:48 51:19:50 280.5 1964 112.21
Marie 2400 p1141_RIBO_FSpeptide_HEL_nospring Gromac (Big) 600 100 0:32:04 53:26:40 269.4 1886 112.27
WS19-Matt 2500 p1140_RIBO_FSpeptide_EXT_nospring Gromac (Big) 600 100 0:30:47 51:18:20 280.7 1965 112.27
Marie 2400 p1141 checkpoint 30 Gromac (Big) 600 100 0:32:02 53:23:40 269.7 1888 112.37
WS23-Ben 2300 p1141 checkpoint 15 Gromac (Big) 600 100 0:33:25 55:41:40 258.6 1810 112.41
WS23-Ben 2300 p1141 checkpoint 30 Gromac (Big) 600 100 0:33:16 55:26:40 259.7 1818 112.92
Philip 2600 p1141_RIBO_FSpeptide_HEL_nospring Gromac (Big) 600 100 0:29:21 48:54:30 294.4 2061 113.24
WS19-Matt 2500 p1129_L939_K12M_503K_600V Tinker 231 400 0:02:55 19:23:24 285.9 2001 114.37
WS19-Matt 2500 p1129_L939_K12M_503K_600V Tinker 231 400 0:02:54 19:23:07 286.0 2002 114.40
ChasR 2532 p1321_1fnt_a16_9.1ps-1_uf Gromac (Big) 346 100 0:16:29 27:29:00 302.1 2115 119.33
ChasR 2532 p1320_1fnt_a12_9.1ps-1_uf Gromac (Big) 343 100 0:16:19 27:11:50 302.7 2119 119.54
Philip 2600 p1316_1fnt_a6_9.1ps-1_uf Gromac (Big) 343 100 0:15:41 26:07:50 315.0 2205 121.17
WS19-Matt 2500 p1479_tet_1479_1 Gromac (Big) 364 100 0:17:15 28:45:15 303.8 2127 121.53
WS19-Matt 2500 p1479_tet_1479_1 Gromac (Big) 364 100 0:17:14 28:42:55 304.2 2130 121.69
WS19-Matt 2500 p1317 Gromac (Big) 343 100 0:16:14 27:03:15 304.3 2130 121.71
Marie 2400 p1479_tet_1479_1 Gromac (Big) 364 100 0:17:54 29:49:30 292.9 2050 122.05
Marie 2400 p1476_tet_2nm Gromac (Big) 364 100 0:17:49 29:42:20 294.1 2059 122.54
Philip 2600 p1479_tet_1479_1 Gromac (Big) 364 100 0:16:20 27:13:09 321.0 2247 123.44
WS19-Matt 2500 p1476_tet_2nm Gromac (Big) 364 100 0:16:55 28:12:15 309.7 2168 123.90
Philip 2600 p1479_tet_1479_1 Gromac (Big) 364 100 0:16:15 27:05:30 322.5 2257 124.02
ChasR 2532 p1478 Gromac (Big) 364 100 0:16:14 27:03:59 322.8 2259 127.47

With the exception of two no longer current Tinkers, p590 and p1129, the Tinkers don't come close to Big WUs in production on my AMD rigs. The only reason I can think of for you to come to a different conclusion is that assembly optimizations are turned off on your rigs. Tinkers don't use SSE but Gromacs do and go very slowly if SSE isn't working. Another possibility would be lack of memory causing swapping of data to the HD.

[stratcatprowlin]

Yeah on my 3200 Newcastle(Now Frank's)@200x11 it would take me about 30 hours to finish a 241 point tinker and 50-53 hours on a 600 pointer.So the big packets were definitely more productive on that rig.I got Similar results on my XP2800(Now icesaber's lol).

[TollhouseFrank]

hmm... maybe my SSE extensions are turned off or something... i dunno.... but i know the math ChasR shows is flawless... yet it seems my machine seems to defy all that. Its freaky in its own sorta way.

[Mark620]

I am getting 99 hours for p1144_RIBO_nopeptide on my XP2400 with 1 gig ram.

I am getting 33-36 hours for a p1154_L939_K12M_298K_DT_250ps on the same machine.

[stratcatprowlin]

I am getting 99 hours for p1144_RIBO_nopeptide on my XP2400 with 1 gig ram.

I am getting 33-36 hours for a p1154_L939_K12M_298K_DT_250ps on the same machine.

Hmmm.Maybe the extra L2 cache on the Barton core makes that much of a difference on that work unit? Weird.

[Mark620]

If someone is interested in working with me to get "more" out of my machines....just let me know.

I am running windows XP,SP1 on that machine with wedos one click w/fah502....

[ChasR]

If someone is interested in working with me to get "more" out of my machines....just let me know.

I am running windows XP,SP1 on that machine with wedos one click w/fah502....

Mark,
Your Tinker numbers are exactly where they should be. 80-90ppd/GHz so the question is why are the big packets folding so slow? Is it slow on all Gromacs or just the big ones? First thing to check is the log from a Gromac and make sure it says Extra SSE boost OK when it starts the WU (-verbosity 9 is required to see this). If not, that's the problem. The registry should contain the -forceasm flag (I think all the one clicks set this falg) to force SSE on at all times. If SSE is working, it's either a memory/cache latency problem, or thermal throttling problem. I have an extensive database on XPs other than my own and in it, all XPs ,Palaminos, T'breds, Bartons, produce more on big WUs than tinkers so we'll just have to look for the issue with yours.

[Mark620]

Partial log from machine in question:


# Windows Console Edition ################################################## ###
################################################## #############################

Folding@Home Client Version 5.02

http://folding.stanford.edu

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

Launch directory: C:\Program Files\FAH
Executable: C:\Program Files\FAH\FAH502-Console.exe
Arguments: -local -service -verbosity 9 -forceasm

Warning:
By using the -forceasm flag, you are overriding
safeguards in the program. If you did not intend to
do this, please restart the program without -forceasm.
If work units are not completing fully (and particularly
if your machine is overclocked), then please discontinue
use of the flag.

[19:50:08] Folding@Home Gromacs Core
[19:50:08] Version 1.80 (March 16, 2005)
[19:50:08]
[19:50:08] Preparing to commence simulation
[19:50:08] - Assembly optimizations manually forced on.
[19:50:08] - Not checking prior termination.
[19:50:18] - Expanded 3032027 -> 16523193 (decompressed 544.9 percent)
[19:50:18] - Starting from initial work packet
[19:50:18]
[19:50:18] Project: 1144 (Run 87, Clone 10, Gen 1)
[19:50:18]
[19:50:19] Assembly optimizations on if available.
[19:50:19] Entering M.D.
[19:50:27] Protein: p1144_RIBO_nopeptide
[19:50:27]
[19:50:28] Writing local files
[19:50:40] Extra SSE boost OK.
[19:50:42] Writing local files
[19:50:42] Completed 0 out of 250000 steps (0)
[20:20:43] Timered checkpoint triggered.
[20:45:26] Writing local files
[20:45:26] Completed 2500 out of 250000 steps (1)
[21:15:28] Timered checkpoint triggered.
[21:22:15] - Autosending finished units...
[21:22:15] Trying to send all finished work units
[21:22:15] + No unsent completed units remaining.
[21:22:15] - Autosend completed
[21:40:13] Writing local files
[21:40:13] Completed 5000 out of 250000 steps (2)
[22:10:16] Timered checkpoint triggered.
[22:35:01] Writing local files
[22:35:01] Completed 7500 out of 250000 steps (3)
[23:05:04] Timered checkpoint triggered.
[23:29:50] Writing local files
[23:29:51] Completed 10000 out of 250000 steps (4)
[23:59:51] Timered checkpoint triggered.
[00:24:39] Writing local files
[00:24:39] Completed 12500 out of 250000 steps (5)
[00:54:40] Timered checkpoint triggered.
[01:19:36] Writing local files

[ChasR]

SSE is on. Frame times consistent. Do you have frame times from a regular gromac?

[Mark620]

Screen shot of FAH part registry

[Mark620]

Chasr, Found one....Fahlog-prev.txt...same machine....XP2400, winxp, 1gb ram

[03:05:18] *------------------------------*
[03:05:18] Folding@Home Gromacs Core
[03:05:18] Version 1.80 (March 16, 2005)
[03:05:18]
[03:05:18] Preparing to commence simulation
[03:05:18] - Assembly optimizations manually forced on.
[03:05:18] - Not checking prior termination.
[03:05:23] - Expanded 862301 -> 11085361 (decompressed 1285.5 percent)
[03:05:23] - Starting from initial work packet
[03:05:23]
[03:05:23] Project: 1323 (Run 0, Clone 181, Gen 39)
[03:05:23]
[03:05:23] Assembly optimizations on if available.
[03:05:23] Entering M.D.
[03:05:30] Protein: p1323_1fnt_a25_9.1ps-1_uf
[03:05:30]
[03:05:30] Writing local files
[03:05:31] Extra SSE boost OK.
[03:05:31] Writing local files
[03:05:31] Completed 0 out of 2000000 steps (0)
[03:31:07] Writing local files
[03:31:08] Completed 20000 out of 2000000 steps (1)
[03:57:53] Writing local files
[03:57:54] Completed 40000 out of 2000000 steps (2)
[04:23:34] Writing local files
[04:23:34] Completed 60000 out of 2000000 steps (3)
[04:53:35] Timered checkpoint triggered.
[04:56:17] Writing local files
[04:56:18] Completed 80000 out of 2000000 steps (4)
[05:23:24] Writing local files
[05:23:24] Completed 100000 out of 2000000 steps (5)
[05:39:05] - Autosending finished units...
[05:39:05] Trying to send all finished work units
[05:39:05] + No unsent completed units remaining.
[05:39:05] - Autosend completed
[05:48:53] Writing local files
[05:48:53] Completed 120000 out of 2000000 steps (6)
[06:18:54] Timered checkpoint triggered.
[06:47:23] Writing local files
[06:47:24] Completed 140000 out of 2000000 steps (7)
[07:17:14] Writing local files
[07:17:14] Completed 160000 out of 2000000 steps (8)
[07:47:15] Timered checkpoint triggered.
[07:55:07] Writing local files
[07:55:07] Completed 180000 out of 2000000 steps (9)
[08:20:36] Writing local files
[08:20:36] Completed 200000 out of 2000000 steps (10)
[08:46:04] Writing local files
[08:46:04] Completed 220000 out of 2000000 steps (11)
[09:12:14] Writing local files
[09:12:14] Completed 240000 out of 2000000 steps (12)
[09:38:38] Writing local files
[09:38:39] Completed 260000 out of 2000000 steps (13)
[10:04:06] Writing local files
[10:04:06] Completed 280000 out of 2000000 steps (14)
[10:29:36] Writing local files
[10:29:36] Completed 300000 out of 2000000 steps (15)
[10:55:04] Writing local files
[10:55:04] Completed 320000 out of 2000000 steps (16)
[11:21:12] Writing local files
[11:21:12] Completed 340000 out of 2000000 steps (17)
[11:39:05] - Autosending finished units...
[11:39:05] Trying to send all finished work units
[11:39:05] + No unsent completed units remaining.
[11:39:05] - Autosend completed
[11:47:04] Writing local files
[11:47:04] Completed 360000 out of 2000000 steps (18)
[12:13:18] Writing local files
[12:13:18] Completed 380000 out of 2000000 steps (19)
[12:38:32] Writing local files
[12:38:32] Completed 400000 out of 2000000 steps (20)
[13:03:47] Writing local files
[13:03:47] Completed 420000 out of 2000000 steps (21)
[13:29:01] Writing local files
[13:29:01] Completed 440000 out of 2000000 steps (22)
[13:54:17] Writing local files
[13:54:17] Completed 460000 out of 2000000 steps (23)
[14:19:32] Writing local files
[14:19:32] Completed 480000 out of 2000000 steps (24)
[14:44:46] Writing local files
[14:44:47] Completed 500000 out of 2000000 steps (25)
[15:10:01] Writing local files
[15:10:01] Completed 520000 out of 2000000 steps (26)
[15:35:13] Writing local files
[15:35:14] Completed 540000 out of 2000000 steps (27)
[16:05:14] Timered checkpoint triggered.
[16:05:37] Writing local files
[16:05:38] Completed 560000 out of 2000000 steps (28)
[16:30:51] Writing local files
[16:30:51] Completed 580000 out of 2000000 steps (29)
[16:56:03] Writing local files
[16:56:03] Completed 600000 out of 2000000 steps (30)
[17:21:15] Writing local files
[17:21:15] Completed 620000 out of 2000000 steps (31)
[17:39:05] - Autosending finished units...
[17:39:05] Trying to send all finished work units
[17:39:05] + No unsent completed units remaining.
[17:39:05] - Autosend completed
[17:49:09] Writing local files
[17:49:10] Completed 640000 out of 2000000 steps (32)
[18:17:27] Writing local files

[Mark620]

Just installed MBM5. CPU temp is 41C

[ChasR]

Actually p1323 is a 308 pt. big WU. Frame times indicate you were doing something else or something else was sapping cpu cycles on the rig during this snippet. Note that the timered checkpoint occurs between some frames but not all. The difference in frame times is about 13 minutes on a WU that should be consistent. The big WUs are going to be more sensitive to other use of the FPU than a Tinker so a competing process is a possibility.

[Mark620]

This machine is the one my wife uses and she may have left a game running....
Not to mention NAV & all the other uses...

[ChasR]

Mark,
A frame time from a regular Gromac WU would be helpful. If all Gromacs fold slow it's likely a processor related issue. If only the big WUs fold slowly it's probably a memory latency or memory bandwidth issue. It's also possible that even though the core reports SSE OK that there are problems with SSE. SiSoft Sandra would give you a good idea of what's going on in terms of bandwidth, fpu performance, cache performance, and sse performance. But problems of these types would be specific to one rig and you're experiencing TInkers folding faster than big WUs on "AMD machines".
You wouldn't be running something like a 3D clock screensaver on your AMD rigs would you?

[Mark620]

ChasR,

I will get this sometime over the weekend of the 4th.
What machines should I benchmark against in SiSoft Sandra
It is a Gigabyte GA-7VKMP Mobo, XP 2400 Thorton, 1GB PC 2100 (mismatched 2x256)

[ChasR]

ChasR,

I will get this sometime over the weekend of the 4th.
What machines should I benchmark against in SiSoft Sandra
It is a Gigabyte GA-7VKMP Mobo, XP 2400 Thorton, 1GB PC 2100 (mismatched 2x256)

We can compare it to one of my nForce2/XP Bartons. But if Sandra has an nforce 2/XP combination in its db that'll do. Maybe the mismatched ram is causing a reduction in bandwidth. All of my XPs are running dual channel. The ones running on a 333MHz FSB aren't apperciably slower than those on a 400 MHz FSB. I don't have any 333s nearby but I'd be glad to look at your results and compare them to one at hand.

[Mark620]

Man I cant believe I fudged that...memory 2 sticks of 512meg(mismatched)...not as stated above