Large SMP points off

Kendrak

[H]ard|DCer of the Year 2009
Joined
Aug 29, 2001
Messages
21,141
I have seen around here and other places that some of the larger SMP WU are off on their points.

I would like your help to see if that is the case.

PG loves to see a bunch of data before they do anything (I support this). So [H] lets give them some data and fix this if it is off!

I will switch a SMP rig on myself to help with this, but I know there are some X6, 2500k, 2600k, 920/30 and the like out there folding SMP.

Give me your DATA!

I need comparison of a "normal" SMP and a large SMP.
 
Last edited:
low point running on i7 @4.2g
7809 (8. 288, 43) TPF 8:07 ETA ~14hours gets 12939 points.

60985 atoms preferred deadline: 25.6 day, final 55.06day
 
Normal SMP on dual x5450@3GHz:

7131 (Run 0, Clone 81, Gen 415)
TPF ~2:55
ETA ~4.8 hours
4672 Points
23067 PPD
 
Just to clarify, I think you're talking about the 10127 WUs when you say "large". Those are the WUs that give notoriously low PPD and to add insult to injury, they have a large atom count (they take much longer to finish).

I folded a few of them on my laptop (i5 540M, 2.5GHz, dual-core with HT) and my TPF was around 75 minutes which is more than 5 days (!) to fold one unit and all you get is 2300 PPD. Compare that to a 7200 WU, with 8 min TPF and 3500 PPD.
 
Disclaimer: I just decided to take the plunge and start folding this weekend

I'm using the SMP client on my 2600k box and I'd be glad to help out with the data collection, just let me know what I need to look at. I'm still getting my bearings with the acronyms and important numbers.

Also, if anyone could help me optimize my config (-smp -bigadv ?) please let me know. I'm also looking for what I need to edit so I don't have to authorize the program to upload/download WU's.
 
Disclaimer: I just decided to take the plunge and start folding this weekend

I'm using the SMP client on my 2600k box and I'd be glad to help out with the data collection, just let me know what I need to look at. I'm still getting my bearings with the acronyms and important numbers.

Also, if anyone could help me optimize my config (-smp -bigadv ?) please let me know. I'm also looking for what I need to edit so I don't have to authorize the program to upload/download WU's.

Welcome!

Start a thread with all your hardware specs and questioms and we will get you up and running!
 
Just to clarify, I think you're talking about the 10127 WUs when you say "large". Those are the WUs that give notoriously low PPD and to add insult to injury, they have a large atom count (they take much longer to finish).

I folded a few of them on my laptop (i5 540M, 2.5GHz, dual-core with HT) and my TPF was around 75 minutes which is more than 5 days (!) to fold one unit and all you get is 2300 PPD. Compare that to a 7200 WU, with 8 min TPF and 3500 PPD.



I agree on that one I did a few like that but dont have data on them now...
there are a few in that 10xxx wu group like that
 
Kendrak - I have a huge pile of data on my 2600k folding SMP - is there an easy way to extract from HFM.NET that you know of, or are there specific WU's that you'd like compared?

I can get the data that is needed on the work unit history viewer, but i can't seem to export it...
 
K, I have a load of data relating to 6xxx and 7xxx SMP but no data for 10xxx SMP. Let me know if you want anything.

The rig is dual 5640's on windows 7, alas I no longer have my x6 rig but on a good day it would only just do 10.5k on A3 SMP on XP
 
Welcome!

Start a thread with all your hardware specs and questioms and we will get you up and running!

Alright I emailed my log file to myself, but it looks like with the way I am configured now I am getting 6xxx and 7xxx WU's. If that is what you're looking for, let me know what information you'd like.
 
I have ~30k records in HFM...most of them SMP.

What particular units do you need numbers for?
 
I have to check when I get home.

I have been folding a lot of the 10xxx wu's over the last couple weeks.
 
kendrak, Give us Wu numbers. Well give you all of the datas
 
I would really like a person that fold a lot of SMP to give the WU, because my SMP rigs are borgs.

I've got a rig I will turn on..... but as much as I've seen is mentions of large SMP that give bad points.
 
Both my sig rigs run SMP. Let me know how I can get data for you and I'll be happy to oblige.
 
965BE @ 3.7 GHz running a P10127 right now: tpf = 50:49, ppd = 4268, roughly 2.5 days total for completion.

1090T @ 4.0 GHz P7200: tpf = 03:26, ppd = 12470
Q6600 @ 3.0 GHz P7200: tpf = 5:23, ppd = 6350

E7200 @ 3.1 GHz P11021: tpf = 25:54, ppd = 600

(all PCs except E7200 also GPU folding on NVidia hardware)
 
Last edited:
Here's some numbers for the machines I run. All stock speeds. All machines run Windows 7 x64.

Q9650

10127 (80,2,11); 38:23
10128 (73,1,30); 13:11
10129 (59,2,29); 15:01
10130 (47,0,7); 16:03
10131 (22,0,21); 13:06
10132 (75,1,4); 13:02
10133 (52,1,10); 15:20
10134 (28,0,27); 14:14


i7-2600(non k)

10127 (9,0,17); 21:39
10128 (76,1,11); 8:19
10129 (75,0,6); 9:25
10130 (10,0,21); 9:41
10131 (59,2,21); 8:04
10132 (75,2,27); 8:08
10133 (82,2,18); 9:36
10134 (69,0,30); 8:54
 
Just to clarify, I think you're talking about the 10127 WUs when you say "large". Those are the WUs that give notoriously low PPD and to add insult to injury, they have a large atom count (they take much longer to finish).

I folded a few of them on my laptop (i5 540M, 2.5GHz, dual-core with HT) and my TPF was around 75 minutes which is more than 5 days (!) to fold one unit and all you get is 2300 PPD. Compare that to a 7200 WU, with 8 min TPF and 3500 PPD.

Except, Stanford is re calibrating their SMP PPD so many of the new WUs are the low-output kind unless you're either very OC'd or have a Sandybridge.

Put it this way, my sig rig at stock and my [email protected] got about the same PPD. Last September I did about 800k pts in one month in 200WUs. In December I did about 200WUs again, and I only pulled in 300k pts.

Or another way, I used to hit about 15-20k PPD on my rigs with 6050WUs...Now with WUs: 1012x, 700x, 70xx, and 80xx...I am lucky to break 10k ppd.
 
These are on 2600k at 4.7 -smp, the last is a (normal wu)

Project ID: 10126
Core: GRO-A3
Credit: 1208
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:05:47 - 25,056.0 PPD
Avg. Time / Frame : 00:06:25 - 21,439.5 PPD

Project ID: 10127
Core: GRO-A3
Credit: 3262
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:15:43 - 24,819.8 PPD
Avg. Time / Frame : 00:17:35 - 20,974.3 PPD

Project ID: 7130
Core: GRO-A3
Credit: 585
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:02:15 - 34,045.3 PPD
Avg. Time / Frame : 00:02:17 - 31,762.6 PPD
 
8P 8431 48C 48T 2.4Ghz
P6097 TPF: 00:02:00 - 190,712.93 PPD
P7148 TPF: 00:00:42 - 196,193.13 PPD
 
Last edited:
1090T @ 3.7 GHz
WU:7200
tpf = 05:01,
ppd = 11488
 
This is exactly what I am looking for.

These are on 2600k at 4.7 -smp, the last is a (normal wu)

Project ID: 10126
Core: GRO-A3
Credit: 1208
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:05:47 - 25,056.0 PPD
Avg. Time / Frame : 00:06:25 - 21,439.5 PPD

Project ID: 10127
Core: GRO-A3
Credit: 3262
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:15:43 - 24,819.8 PPD
Avg. Time / Frame : 00:17:35 - 20,974.3 PPD

Project ID: 7130
Core: GRO-A3
Credit: 585
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:02:15 - 34,045.3 PPD
Avg. Time / Frame : 00:02:17 - 31,762.6 PPD
 
These are on 2600k at 4.7 -smp, the last is a (normal wu)

Project ID: 10126
Core: GRO-A3
Credit: 1208
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:05:47 - 25,056.0 PPD
Avg. Time / Frame : 00:06:25 - 21,439.5 PPD

Project ID: 10127
Core: GRO-A3
Credit: 3262
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:15:43 - 24,819.8 PPD
Avg. Time / Frame : 00:17:35 - 20,974.3 PPD

Project ID: 7130
Core: GRO-A3
Credit: 585
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:02:15 - 34,045.3 PPD
Avg. Time / Frame : 00:02:17 - 31,762.6 PPD

Yep, this shows the lack of balance with 10126 and 10127 very nicely. I guess around 1.5x increase in base points would do the trick, as it is the same ratio I had observed with my laptop on these units (~3500PPD with 7200 vs ~2300PPD with 10127). Especially 10127 is terrible because it is really a big unit compared to standard SMP unit.
 
Here's my data. It's a small sample size, but so far my rig has been really consistent.

Project ID: 6972
Core: GRO-A3
Credit: 552
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:02:22 - 30,099.8 PPD
Avg. Time / Frame : 00:02:23 - 29,784.7 PPD

Project ID: 6990
Core: GRO-A3
Credit: 552
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:02:19 - 31,220.5 PPD
Avg. Time / Frame : 00:02:21 - 30,558.6 PPD

vs

Project ID: 10132
Core: GRO-A3
Credit: 1145.36
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:05:50 - 22,838.1 PPD
Avg. Time / Frame : 00:05:54 - 22,452.1 PPD

Project ID: 10134
Core: GRO-A3
Credit: 1264.12
Frames: 100
Number of Frames Observed: 291
Min. Time / Frame : 00:06:30 - 22,509.0 PPD
Avg. Time / Frame : 00:06:40 - 21,670.2 PPD
 
Last edited:
I've made a thread in the DAB.

More than likely more data will be needed to plot a curve for different systems that will be folding these.
 
Adding a new WU

Project ID: 10128
Core: GRO-A3
Credit: 1200.79
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:06:07 - 22,836.8 PPD
Avg. Time / Frame : 00:06:15 - 22,109.9 PPD
 
Adding a new WU

Project ID: 10128
Core: GRO-A3
Credit: 1200.79
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:06:07 - 22,836.8 PPD
Avg. Time / Frame : 00:06:15 - 22,109.9 PPD

Can you give you CPU and speed?
 
Let me know if you want me to stop posting, but here's a new one.

Project ID: 10129
Core: GRO-A3
Credit: 1359.62
Frames: 100
Number of Frames Observed: 100
Min. Time / Frame : 00:06:51 - 23,214.1 PPD
Avg. Time / Frame : 00:07:08 - 21,844.8 PPD

Project ID: 10133
Core: GRO-A3
Credit: 1357.17
Frames: 100
Number of Frames Observed: 125
Min. Time / Frame : 00:06:47 - 23,492.2 PPD
Avg. Time / Frame : 00:07:00 - 22,410.0 PPD
 
Last edited:
100 more frames on the 10133

Project ID: 10133
Core: GRO-A3
Credit: 1357.17
Frames: 100
Number of Frames Observed: 225
Min. Time / Frame : 00:06:47 - 23,492.2 PPD
Avg. Time / Frame : 00:07:02 - 22,250.9 PPD
 
Back
Top