stephenbrooks.orgForumMuon1GeneralBenchmarking with Mpts thread
Username: Password:
Page 1 2 3 4 5 6 7 8 9 10 11 12 13
Search site:
Subscribe to thread via RSS
[XS]riptide
2007-07-18 02:13:03
I'll have another go again.  Got new RAM.  2 x 1GB PC6400 3-3-3-8 Team Xtreem.  The chip benches on current cooling up to 3.85Ghz. 
Stephen, QX6700's have gone beyond 5Ghz.... http://www.xtremesystems.org/forums/showthread.php?t=145209

[XS]riptide
2007-07-18 02:18:07
>Assimilator.....

As for my E6600 score.... I said "E6600 @ 3.6. 1 instance of DPAD at roughly 90-95% CPUusage with 5-10% running 2 x instances of Seventeenorbust."

Notice the Seventeen or bust.... thats massive RAM hampering going on there.  I'll give the E6600 a go too.  That bTW benches at 4.0Ghz, at current cooling. 
[TA]Assimilator1
2007-07-20 19:12:32
lol you benchmarked DPAD with another CPU intensive program?  that messes up the score.
Also that means your gonna whip my arse when you post real scores! 4GHz is awesome!:Q
Stephen Brooks
2007-07-23 11:51:43
So do you two have any "final" scores yet so I can plot the graph?
[XS]riptide
2007-07-24 14:46:57
Sorry Stephen.  I'll get on this in about.... 6 hours.  leave it overnight with the stats recorder on and see what i can do.  results then tomorrow hopefully at 6:15 am UK time.

Assimilator.  The e6600 has been 'retired' to a damaged S3 board where it will live out the rest of its useful life on AIR.  It'll be doing DPAD and SOB at different points.  I don't have enough WC gear around for everything, so the Quads naturally take priority on the cooling.  However... Should you make a shot at it... i can change around stuff.... so go for it.
[XS]riptide
2007-07-24 14:48:25
As for CPU intensive... SOB basically used about 5%. However it would have a dramatic effect on the memory bandwith available for anyhting else.  And i suspect DPAD is a memory BW hog.  Would you agree Stephen?
Stephen Brooks
2007-07-24 15:14:15
Sometimes it's looked like memory latency rather than bandwidth is the key thing, though I suspect it uses quite a bit of both as the particles do not fit inside cache.  Can't remember exactly right now though - I think I examined one of the previous benchmark charts at one point and found a strong correlation with fast FSBs, and AMD generally doing a lot better than average.
HaloJones
2007-07-24 21:51:22
riptide, if you need some more water-cooling how about heading over to Gloucester!  Where are you in the good old UK?
[XS]riptide
2007-07-25 08:43:52
I'm in Ireland.  We've got our own fair share of rain this summer.  50 days and counting where we had rain.  Broke all previous records i believe.  Although you guys got REALLY f*cked.  I spent a summer in the Midlands with my dad actually in Warwickshire, not far form lemmington Spa, Rugby, Warwick.  Lovely country, nice people. 

Anyway back to results....

Uptime (secs),Mpts in file,Estimate kpts/sec
561,499457.6,0.00
862,499551.4,311.46
1163,499655.9,329.22
1464,499817.2,398.01
1765,499838.8,316.44
2067,499977.1,344.99
3572,500539.8,359.35
3874,500644.5,358.28
4175,500751.5,358.04
4476,500842.9,353.84
4777,500975.8,360.08
5078,501082.1,359.61
5379,501138.4,348.82
5681,501264.2,352.87
5982,501366.3,352.10
6283,501449.4,348.09
6584,501580.4,352.43
6885,501666.2,349.22
7186,501773.8,349.58
7488,501875.2,349.02
7789,501976.2,348.45
8090,502004.1,338.22
8391,502110.9,338.85
8692,502217.5,339.41
10198,502767.4,343.44
10499,502888.9,345.26
10800,503014.6,347.38
11102,503084.3,344.07
11403,503225.6,347.54
11704,503254.2,340.71
12005,503440.1,347.99
12306,503466.0,341.27
12607,503597.5,343.66
12909,503728.5,345.88
13210,503834.8,346.05
13511,503913.3,344.07
13812,503941.1,338.35
14113,504098.6,342.45
14716,504309.3,342.76
15017,504412.6,342.77
15318,504497.4,341.52
15619,504630.2,343.51
15920,504696.1,341.06
16221,504775.0,339.54

Average of 346.17, not including '0' at the top.

Short run I think 4 hours. 

That is QX6700 @ 3.69ghz.  410FSB x 9. Ram - 3-3-3-8 1 x 1GB Team Xtreem PC6400 3-3-3-8. Temps were approx 58-62C across 4 cores.  This again, was done with 4 Muon1.exe instances pegged to each core.  Total CPU time was within a minute of each other between the instances.

[XS]riptide
2007-07-25 08:48:53
346.17 x 4 = 1384.68.

Last score was 1273.87. So a 3% increase in clocks and going from 5-5-5-15 to 3-3-3-8 gives me 8% increase in score.
[XS]riptide
2007-07-25 08:52:16
That should read 2 x 1GB Team Xtreem. 
Stephen Brooks
2007-07-25 12:14:51
That's quite an improvement and certainly suggests RAM latency/timings have a lot to do with it.  (When you have a 'final' set up, try running a muonbench on every instance.  I like to be able to add those up rather than extrapolating).
[XS]riptide
2007-07-25 15:20:17
Ok.  I'll do that tonight.
peterzal
2007-08-13 10:28:31
I just started running the muon client today (using wine under linux).  I'm not able to run the benchmark program because it crashes before it can produce any results.  I'm not even sure if that benchmarking program even works under wine.  I would like to get an idea of how fast this client is running.  Would taking the total Mpts and dividing it by the total run time be a good way to measure the speed of the client?  Im running this on an Athlon XP 2500+ (Barton) @ 1.83Ghz.  My client got 73.5 Mpts in 480 seconds (2 auto-saves).  This comes out to 0.153 Mpts/sec.  Would mutliplying this value by 1000 give me an estimate of the number of kpts/sec I'm getting? 
[OCAU] badger
2007-08-13 13:17:02
I've got the benchmark program running under wine, I can't remember what I did to get it working though, I'll look at it tomorrow
[OCAU] badger
2007-08-13 13:18:45
Oh, and you have the idea for a manual speed test, but you should run it for longer than one run.  Try running it for 12 - 24 hours and calculate the Mpts produced in that time.  I use zeonX's muoncalc to add up the mpts in a results file.
peterzal
2007-08-13 20:55:59
where can i find this muoncalc utility? 
Stephen Brooks
2007-08-15 14:55:54
Odd, the hosting for ZeonX's (mega-reviews) site that used to have it seems to have stopped working.  I've sent him and instant message asking where it might have gone.
Stephen Brooks
2007-08-15 15:17:25
He fixed it!

http://muon.mega-reviews.com/muoncalc.asp

And a nice site here:

http://muon.mega-reviews.com/
[TA]Assimilator1
2007-08-16 20:38:22
Well I finally got around to doing a 'final' benchmark with my E6420 @3.2GHz ,400/400MHz FSB/RAM 4-4-4-15 ,but the final score showed me it wasn't quite as fast as 3.2GHz 400/426MHz FSB/RAM 4-4-4-15 ,despite earlier tests showing otherwise!.
So now I'm confused ,earlier tests showed that with the RAM synced @400MHz I was getting about 292 Kpts/s ,& at 426MHz I was getting 289 Kpts/s.Ok I know it's a tiny difference but I'll take any speed boost which is free ,anyway the final score (which I benchmarked for about 20hrs IIRC) gives about 287 Ktp/s (figures below),wth?? 
Is this a variance within the benchmark?  or quite possibly I may of forgotten to turn off sample d/ls on the shorter tests,does that make much difference?
Btw it was definetly off for this 'final' score.

I think I'm going to do a full benchmark run with the RAM at 426MHz too.

Riptide
Did you run the benchmark?

*********************************************************************
Final run for C2D E6420 @3.2GHz, 400/400 FSB/RAM 4-4-4-15 ,2 clients set to single threads.

Core 0
Uptime (secs),Mpts in file,Estimate kpts/sec
24185,2351571.1,0.00
24487,2351618.3,156.56
25090,2351817.5,272.44
25391,2351880.2,256.33
25693,2351999.1,283.94
25994,2352048.4,263.87
26296,2352160.6,279.34
27200,2352361.1,262.05
27803,2352559.1,273.11
31119,2353564.3,287.46
31421,2353640.2,285.97
31722,2353683.7,280.31
32024,2353790.5,283.15
35641,2354796.8,281.58
36244,2354996.9,284.09
38355,2355646.8,287.65
38957,2355766.8,284.03
41068,2356412.9,286.80
41671,2356586.4,286.83
45288,2357592.5,285.33
48605,2358597.1,287.73
49207,2358705.3,285.12
49509,2358825.7,286.48
52825,2359833.5,288.49
53428,2359988.4,287.84
57046,2360993.6,286.74
59156,2361601.0,286.81
59458,2361711.9,287.50
59759,2361751.4,286.17
60362,2361933.0,286.42
60965,2362153.9,287.73
64583,2363159.0,286.85
65186,2363315.7,286.45
68502,2364320.1,287.68
68803,2364440.1,288.43
69406,2364558.7,287.20
69708,2364668.7,287.72
73326,2365674.7,287.01
76642,2366680.5,288.04
76943,2366718.0,287.10
77245,2366821.0,287.41
77546,2366906.9,287.40
79657,2367512.7,287.38
82973,2368517.4,288.26
83275,2368539.6,287.17
83576,2368648.1,287.54
83877,2368727.0,287.41
85988,2369330.4,287.36
86289,2369438.4,287.70
86591,2369485.2,287.06
89907,2370491.0,287.88
90209,2370589.3,288.05
90510,2370622.5,287.24
93827,2371627.3,287.99

Core 1

Uptime (secs),Mpts in file,Estimate kpts/sec
24809,1432432.7,0.00
25712,1432655.3,246.53
26314,1432849.8,277.16
26916,1433046.7,291.43
27819,1433244.7,269.78
28421,1433433.5,277.09
29023,1433632.9,284.83
29625,1433833.3,290.84
30227,1433951.4,280.32
32334,1434552.1,281.66
35645,1435560.4,288.66
37752,1436166.7,288.51
38053,1436213.6,285.50
41364,1437222.5,289.34
41665,1437237.3,285.05
43772,1437840.4,285.19
44072,1437939.6,285.88
44373,1438044.6,286.85
44674,1438163.7,288.50
46781,1438766.7,288.28
47383,1438898.8,286.44
47985,1439093.7,287.41
51597,1440172.8,288.94
52500,1440422.7,288.55
56112,1441428.8,287.39
59423,1442437.1,289.03
61530,1443039.2,288.85
61831,1443055.7,286.95
62132,1443159.3,287.41
62432,1443275.3,288.19
62733,1443390.7,288.95
63034,1443411.8,287.22
63335,1443487.6,286.95
65442,1444091.5,286.93
65743,1444241.4,288.48
66044,1444291.9,287.60
68151,1444897.2,287.59
71763,1445904.3,286.91
72064,1446021.8,287.57
72666,1446208.2,287.85
74773,1446812.1,287.80
75074,1446930.0,288.42
75375,1447005.9,288.21
76278,1447204.2,287.00
79589,1448210.2,288.02
79890,1448330.3,288.63
80793,1448527.5,287.49
81395,1448752.3,288.41
81696,1448830.8,288.26
85307,1449849.8,287.90
85909,1450025.7,287.94
88016,1450631.3,287.92
90123,1451234.2,287.86
92230,1451836.8,287.81
92531,1451904.6,287.53
93133,1452102.5,287.89

If the average for those is 287 then my rig is now getting 574 Kpt/s
[TA]Assimilator1
2007-09-18 21:13:27
lol ,can't believe 1mth has passed!  :Q

Anyway, here's the benchmarks with my C2D E6420 @3.2GHz, 400/427 FSB/RAM 4-4-4-15 ,2 clients set to single threads.
I wanted to see whether the faster RAM speed was negated by the increased latency running the RAM asynchronously.

Core 0

Uptime (secs),Mpts in file,Estimate kpts/sec
3554,499040.5,0.00
5657,499652.3,291.00
5957,499766.2,302.03
8060,500378.5,296.99
8961,500582.0,285.13
9261,500702.9,291.31
9561,500756.7,285.70
9862,500880.1,291.66
10462,501071.4,293.99
12565,501683.5,293.33
13466,501886.8,287.17
13766,502002.8,290.09
14066,502047.0,286.00
16169,502659.1,286.86
16770,502861.9,289.17
17370,503069.1,291.59
18271,503272.9,287.59
18572,503395.1,289.97
19172,503599.0,291.88
20073,503801.9,288.24
22176,504453.8,290.70
24278,505066.0,290.75
27882,506088.3,289.70
28182,506210.9,291.15
28783,506311.5,288.20
29384,506531.8,290.03
31486,507163.6,290.82
31787,507240.6,290.45
32087,507333.3,290.64
35691,508355.3,289.85
35991,508441.1,289.81
36592,508599.5,289.34
36892,508716.8,290.25
37493,508829.8,288.44
39596,509442.0,288.60
39896,509564.4,289.58
41998,510176.5,289.67
42599,510379.3,290.41
43500,510581.6,288.92
43801,510695.3,289.59
45903,511306.5,289.64
46504,511419.3,288.22
47104,511624.5,288.96
49207,512236.5,289.05
49807,512436.4,289.62
50108,512514.7,289.44
50708,512654.1,288.70
52811,513266.4,288.81
53412,513422.0,288.45
55514,514033.8,288.56
56115,514234.9,289.08
58217,514846.7,289.16
61821,515884.2,289.08
62122,516006.2,289.68
63023,516210.0,288.72
65125,516863.1,289.46
65426,516942.1,289.34
65726,517039.0,289.50
66026,517057.9,288.41
66327,517180.5,288.98
66627,517302.9,289.55
67528,517506.7,288.65
67828,517629.1,289.21
68730,517885.3,289.14
70832,518496.3,289.19
72935,519108.5,289.25
73836,519329.6,288.68
74436,519532.9,289.11
75037,519736.2,289.52
78641,520758.1,289.23
78942,520865.8,289.51
79242,520932.3,289.24
79843,521107.9,289.26
80744,521310.9,288.52

Core 1

Uptime (secs),Mpts in file,Estimate kpts/sec
2630,298884.0,0.00
2930,298977.9,312.69
5032,299589.7,293.77
5633,299795.3,303.48
7735,300427.0,302.26
11338,301449.1,294.56
11638,301571.6,298.34
12239,301680.8,291.06
14341,302299.8,291.67
14641,302422.2,294.57
15542,302626.0,289.80
15842,302748.4,292.48
17944,303359.1,292.21
18245,303482.3,294.48
18545,303523.2,291.50
19145,303695.9,291.35
19446,303818.3,293.43
21548,304428.8,293.10
21848,304505.0,292.48
22449,304627.2,289.79
23049,304821.2,290.76
25151,305433.5,290.81
25451,305555.7,292.34
27854,306226.8,291.11
29956,306838.2,291.09
30256,306876.6,289.31
33559,307898.6,291.46
34460,308131.8,290.54
36562,308743.3,290.56
37163,308865.9,289.06
37463,308953.4,289.08
40766,309975.7,290.84
42868,310586.2,290.82
43769,310830.0,290.38
45871,311441.2,290.40
47973,312051.6,290.40
48273,312101.6,289.58
48874,312285.8,289.81
49475,312448.9,289.57
49775,312571.2,290.32
50375,312729.9,289.99
53979,313752.2,289.55
57583,314774.4,289.17
57883,314897.0,289.81
58483,315101.0,290.35
58784,315121.2,289.16
60886,315733.2,289.23
61186,315801.9,288.92
61787,316005.5,289.43
63889,316640.2,289.86
64189,316662.0,288.80
64790,316865.3,289.28
65090,316985.8,289.81
65991,317217.1,289.34
66291,317327.8,289.72
66591,317398.8,289.47
67192,317564.4,289.34
68093,317767.9,288.47
70195,318378.5,288.53
70495,318542.6,289.67
74099,319564.8,289.37
74399,319664.3,289.54
75000,319836.2,289.52
75300,319863.4,288.69
75600,319985.8,289.18
75901,320033.0,288.64
76201,320124.9,288.71
76801,320290.3,288.61
77402,320493.9,289.01
78003,320678.9,289.16
80105,321290.8,289.21
82207,321903.0,289.27

At a glance ,it looks 289 Kpts/s/core
Total 578 Kpts/s ,that's not even a 1% boost
Probably that's within the margines of error?  (Stephen?)

Certainly not worth the extra power used (& heat output) as a result of the RAM needing an extra 0.1v.

I was going to test @ 3.25GHz ,406/380 FSB/RAM 4-4-4-15 (can't get it stable at 406/406 without using overly high NBv) ,but seeing as the FSB/CPU boost would only be about 1.5% & the RAM would asynch'ed and underclocked I highly doubt it would boost scores ,more likely to drop them.

My next test will be @3.2GHz, 400/400 FSB/RAM 4-4-4-15 Command rate 1T (currently at 2T) ,hopefully that could boost scores by about 5% IF I can get it stable at those settings.....

Btw is no-one else reading this thread anymore? 
[TA]Assimilator1
2007-09-18 21:15:33
Forgot to say ,the above run was for about 22hrs
Stephen Brooks
2007-09-18 23:42:36
It has been just recently.  Maybe I should do another graph now.  I was waiting for a while on certain people who said "these aren't final results".
[XS]riptide
2007-09-25 01:32:25
Ya... that quad is now in another machine.  Never did get a chance to run the benchs again... .
Stephen Brooks
2007-10-01 11:59:10
I think I got this right - all the updates I could find added to the graph (which now goes to 1400 instead of 1100!):

Stephen Brooks
2007-10-02 02:16:46
I was just looking at this again, specifically the pts/Mclock/core figures.  Although the Core 2 quads are winning by brute force right now, AMD's K8/9 architecture still have a 10% advantage on efficiency per clock (100ish vs. 90ish).  It looks jumbled but when you start looking at the distribution of figures for each separately it becomes clearer.  P4 architecture seems to score a consistent 70ish.  Muon1 is a HPC workload and so the usual clock-for-clock advantage of Core 2 over Athlon is reversed.

There's some new architectures coming down the pipe this winter (from both manufacturers), will be interesting to see those in comparison.
[TA]JonB
2007-11-26 14:42:51
I am hoping someone will get a new Phenom processor from AMD and post results.  That is my most likely upgrade, probably mid 2008.
tomaz
2007-11-29 20:06:54
X2 4800+ 2.5 GHz (actually running on 2.487 Mhz because of some MB issues - fsb 199 x 12.5=2487).  Ram is ddr2 800 but also running at 710 or so because of other issues...yap JonB, I'd also like so see Phenom results.  Time for some advanced micro devices to break 1500 margin

Average of a "flat part" is 510.

Uptime (secs),Mpts in file,Estimate kpts/sec
124522,38621.0,0.00
124822,38808.4,624.65
125122,38963.5,570.80
125422,39117.1,551.20
125722,39283.7,552.23
126022,39435.6,543.05
126322,39528.1,503.92
126622,39679.3,503.94
126922,39866.2,518.81
127222,40021.3,518.61
127522,40231.4,536.78
127822,40382.9,533.89
129622,41248.3,515.14
129922,41415.7,517.52
131723,42253.1,504.44
133223,43087.6,513.38
133523,43269.9,516.52
133823,43347.0,508.15
135323,44186.4,515.29
137123,45018.2,507.69
137423,45208.2,510.62
137723,45312.1,506.88
138023,45519.5,510.98
138323,45692.9,512.44
139823,46529.2,516.86
140123,46630.2,513.39
141923,47466.4,508.34
142223,47671.5,511.31
142523,47804.2,510.16
142823,47940.4,509.24
143123,48072.0,508.10
143423,48268.4,510.43
143723,48433.6,511.05
144023,48593.9,511.41
144323,48761.1,512.11
144623,48920.3,512.38
144923,49110.6,514.18
145523,49302.9,508.64
147023,50185.2,513.94
148823,51001.3,509.46
149123,51167.5,510.00
150623,52025.4,513.56
152423,52880.6,511.08
152723,53047.2,511.55
154523,53891.1,508.98
154823,54054.6,509.34
156323,54896.9,511.80
158124,55734.3,509.30
159624,56521.9,509.98
159924,56690.0,510.40
161424,57532.4,512.48
161724,57657.7,511.72
162024,57761.5,510.39
162324,57928.5,510.76
164124,58774.2,508.90
164424,58935.2,509.11
165924,59797.9,511.50
167724,60640.2,509.68
168024,60858.2,511.18
169824,61695.1,509.34
170124,61835.8,509.07
171624,62674.7,510.67
173424,63517.0,509.10
173724,63721.3,510.15
175224,64554.2,511.48
177024,65399.3,510.04
177325,65645.9,511.81
177625,65705.7,510.05
179125,66556.1,511.61
179425,66712.4,511.66
181225,67556.7,510.31
181525,67700.7,510.15
181825,67869.2,510.42
182125,67994.0,509.93
182425,68136.7,509.75
183925,68979.0,511.05
185725,69838.7,510.07
187225,70681.0,511.30
187525,70707.9,509.29
187825,70877.3,509.56
188125,71043.4,509.76
188425,71216.6,510.08
188725,71383.9,510.30
190225,72216.1,511.32
192025,73058.6,510.16
193525,73904.4,511.33
193825,73925.7,509.42
195325,74764.5,510.48
195625,74850.7,509.54
195925,75028.0,509.88
197426,75866.8,510.89
199226,76708.8,509.85
199526,76853.8,509.75
199826,76996.0,509.60
200126,77162.9,509.79
200426,77335.7,510.05
200726,77488.4,510.05
202226,78317.2,510.87
204026,79155.1,509.84
205526,79996.1,510.78
207326,80824.3,509.68
208826,81668.8,510.63
210626,82508.1,509.70
212127,83326.6,510.31
213927,84295.0,510.87
caferace
2007-11-30 01:16:52
[TA]caferace here...

CPU : Intel Core 2 Duo E4500 (*1)
CPU Arch : 2 Cores - 2 Threads
CPU PSN : Intel(R) Core(TM)2 Duo CPU E4500 @ 2.20GHz
CPU EXT : MMX SSE SSE2 SSE3 SSSE3 EM64T
CPU Cache : L1 : 2 x 32/2 x 32 KB - L2 : 2048 KB
Core : Conroe (65 nm) / Revision : M0
CPUID : 6.F.D / Extended : 6.F
Freq : 2554.99 MHz (232.27 * 11)
----------------------------------
MB Brand : ASUSTeK Computer INC.
MB Model : P5L-VM 1394
NB : Intel i945G/GZ rev A2
SB : Intel 82801GB (ICH7/R) rev A1
----------------------------------
RAM Size : 2048 MB
RAM Freq : 387.1 MHz
RAM Type : DDR2-SDRAM Dual Channel
RAM Ratio : 3:5
RAM Timings : 5-5-5-15

===========================

Uptime (secs),Mpts in file,Estimate kpts/sec
1216275,21603.7,0.00
1216575,21692.1,294.65
1220175,23366.3,451.93
1221975,24242.9,462.99
1222275,24267.4,443.93
1225876,26114.9,469.89
1226476,26286.6,459.08
1227076,26637.5,466.06
1227376,26765.6,465.01
1229476,27643.4,457.53
1233076,29438.0,466.30
1235176,30298.3,460.01
1235476,30376.1,456.87
1239076,32157.5,462.86
1239676,32501.3,465.68
1240876,32912.8,459.69
1242976,33875.5,459.59
1246877,35714.9,461.13
1248077,36114.5,456.29
1250177,37062.3,455.98
1250477,37210.9,456.33
1250777,37257.8,453.72
1251077,37431.1,454.79
1251377,37587.0,455.34
1254977,39304.1,457.35
1255577,39586.0,457.54
1255877,39695.6,456.84
1259777,41494.4,457.23
1260077,41605.7,456.64
1260377,41802.8,458.00
1260677,41824.7,455.40
1261277,42185.9,457.36
1261577,42339.6,457.72
1261877,42359.6,455.15
1262177,42520.1,455.67
1262478,42636.1,455.22
1262778,42743.4,454.59
1263078,42866.8,454.32
1266978,44646.2,454.46
1267278,44852.6,455.83
1267578,44962.1,455.30
1267878,45102.5,455.38
1271778,46933.2,456.36
1272679,47289.8,455.40
1272979,47455.2,455.91
1273279,47605.5,456.14
1274479,48049.7,454.37
1274779,48172.0,454.13
1278379,49963.8,456.65
1278679,50031.4,455.54
1282280,51820.7,457.80
1283480,52265.7,456.25
1287381,54089.3,456.86
1287981,54434.0,457.85
1288881,54788.7,457.06
1292481,56567.3,458.80

-jim
[TA]Assimilator1
2007-12-11 20:06:21
Jim
Is that 1 or 2 clients you're using?
[XS]riptide
2007-12-12 20:58:35
C'mon... somebody needs to crank it up!  I wanna see Dual Quads, 8 instances @ ~3.2 Ghz!
[TA]Assimilator1
2007-12-13 00:53:09
lol

*NEW CLIENT RELEASED* ,v4.44 now available.

Notes from 'Version History' :-
'It includes a more accurate collision detection algorithm so simulations may take longer to complete but will also register more Mpts'

For future benchmarking you will need to include the client version so scores go into the right graph.
iNSaNiTi
2008-01-23 01:28:41
Bench :


Core 1 :

Uptime (secs),Mpts in file,Estimate kpts/sec
1299,960025.7,0.00
2501,960457.7,359.44
3703,960887.5,358.59
4904,961319.5,358.89
6106,961749.7,358.68
7307,962180.2,358.61
8509,962608.8,358.29
9410,963037.3,371.32
10611,963461.8,368.99
11813,963892.5,367.79
13014,964320.2,366.57
14216,964749.3,365.70
15417,965165.2,364.04
16318,965594.9,370.81
17520,966025.9,369.91
18721,966456.9,369.14
19923,966886.8,368.41
21124,967288.3,366.33
22025,967716.0,371.04
28033,969870.9,368.27
29234,970301.6,367.85
34942,972454.2,369.43
36143,972884.3,369.03
37345,973314.6,368.67
38547,973746.9,368.38
39748,974177.2,368.06
40649,974612.0,370.68
41851,975044.2,370.36
43052,975470.0,369.90
44254,975919.6,370.02
45456,976350.0,369.69
46657,976735.8,368.41
47558,977166.2,370.54
48759,977596.9,370.23
49961,978026.7,369.92
51162,978456.1,369.62
52364,978903.1,369.68
53565,979332.1,369.39
54767,979763.6,369.16
55968,980193.5,368.91
56869,980624.6,370.68
58070,981052.9,370.39
59272,981480.8,370.09
60473,981906.5,369.77
61674,982332.7,369.48
62875,982762.2,369.24
63776,983191.6,370.79
64977,983623.0,370.57
66179,984054.3,370.36
67380,984471.7,369.94
73086,986625.6,370.54


Core 2 :

Uptime (secs),Mpts in file,Estimate kpts/sec
1335,690727.6,0.00
2536,691157.8,358.14
7941,693308.7,390.68
9143,693740.0,385.81
10344,694161.8,381.19
11245,694592.7,390.02
12446,695022.1,386.50
13647,695449.3,383.49
14548,695877.2,389.73
14848,695894.3,382.34
15749,696323.7,388.23
16950,696751.7,385.77
18152,697180.5,383.72
19052,697611.3,388.52
19353,697634.6,383.34
20254,698062.6,387.71
21455,698493.2,385.96
22656,698925.0,384.47
28061,701079.0,387.31
29262,701508.0,386.01
30163,701939.0,388.90
30464,701974.1,386.09
31365,702403.0,388.79
32566,702830.6,387.53
33767,703261.7,386.47
34668,703691.3,388.91
35869,704120.8,387.82
37071,704554.1,386.91
38272,704985.4,386.00
39173,705418.9,388.27
40374,705848.0,387.31
41575,706280.3,386.49
42476,706710.2,388.48
43677,707141.1,387.63
44879,707571.3,386.82
45780,708001.8,388.66
46981,708431.5,387.85
48182,708851.0,386.86
49083,709281.4,388.57
50284,709712.2,387.84
51486,710141.0,387.10
52687,710571.2,386.42
53588,711001.3,387.99
54789,711430.1,387.29
55990,711861.6,386.68
56891,712291.4,388.14
58092,712722.4,387.52
59294,713152.6,386.91
60195,713528.0,387.37
61396,713957.4,386.77
62297,714385.6,388.08
63498,714817.6,387.53
64699,715247.0,386.96
65600,715674.5,388.19
66801,716100.9,387.58
68002,716531.5,387.05
68903,716961.3,388.25
70105,717393.5,387.76
71306,717825.2,387.27
72207,718254.4,388.40
73408,718707.6,388.22

Core 1 (370,54) + Core 2 (388,22) = 758,76

Setup :

*AMD Opteron 170 @ 3120 Mhz (312x10)*
*Dfi Lanparty Ultra-D Rev A3*
*G Skill 2GBNS @ 195 Mhz 2-2-2-5 1T*

Im running this setting almost 24/7 when its below 20 degrees in my room.
Its a nice score for a (S939) Opteron

Greetz,
iNSANiTi





iNSaNiTi
2008-01-24 00:52:07
Forgotn to add the version, its 443D

Later i will add a 444D bench result with the same setting's.
[TA]Assimilator1
2008-02-26 23:06:38
Yea it will be interesting to see the difference.

About time I benchmarked my C2 quad!
[TA]Assimilator1
2008-03-02 13:19:40
v444d benchmarks, & wow this version certainly outputs far less results than the previous versions!
The below tables are for about a 36hr period (I can't work out the 'uptime' column ,I thought it was in 300s intervals but that doesn't tally up)
Sample d/ls turned off.
1 client/core single threaded.

Core 1

1100747,2221348.5,216.84
1105562,2222392.2,216.80
1107066,2222705.7,215.60
1109474,2223270.6,219.14
1126024,2227119.0,226.66
1131440,2228263.4,224.27
1133547,2228692.8,223.11
1138963,2229829.8,221.42
1141972,2230561.9,222.87
1142273,2230625.5,222.80
1143176,2230828.8,222.84
1145884,2231421.5,222.62
1148593,2232104.2,224.16
1148894,2232162.9,223.99
1152204,2232897.6,223.87
1155514,2233639.6,223.88
1159727,2234556.0,223.46
1162736,2235235.5,223.57
1163037,2235319.5,223.82
1167551,2236272.8,223.01
1172968,2237413.9,222.14
1178083,2238411.3,220.43
1181695,2239144.5,219.70
1184403,2239769.8,220.04
1187713,2240506.3,220.13
1188315,2240666.9,220.44
1191625,2241408.0,220.56
1196741,2242501.4,220.21

Core 2

1086011,2095459.0,199.45
1091729,2096605.8,200.18
1092030,2096651.3,198.55
1096544,2097565.9,199.91
1100757,2098480.9,204.01
1102261,2098878.2,208.70
1102562,2098939.4,208.62
1118210,2102575.4,219.17
1118511,2102598.7,217.97
1118812,2102644.9,217.43
1123326,2103661.0,218.29
1150107,2109395.7,216.63
1153117,2110113.9,217.57
1156427,2110849.9,217.79
1183210,2116393.7,214.90
1203973,2120875.8,215.07
1209089,2121978.5,215.09
1209390,2122000.0,214.74
1211797,2122605.9,215.44
1213302,2122920.2,215.36

Core 3

1079372,895687.8,244.79
1081773,896241.3,237.67
1085675,897051.4,224.21
1090776,898073.6,215.40
1111783,902549.7,214.00
1115084,903278.3,214.58
1115685,903427.5,215.11
1131590,907130.7,220.27
1134891,907868.4,220.45
1136091,908156.6,220.85
1140293,909070.1,220.62
1140593,909131.1,220.54
1143594,909726.2,219.54
1148395,910795.7,219.75
1151997,911593.8,219.84
1154998,912224.1,219.47
1155898,912460.5,219.96
1156798,912606.8,219.31
1161300,913559.0,218.89
1166101,914606.5,218.85
1169402,915358.6,219.18
1186208,919063.4,219.37
1190409,919970.9,219.25
1190710,920007.7,218.99
1207215,923884.4,221.00
1211416,924776.2,220.73

Core 4

1073376,898031.3,192.40
1076677,898801.7,220.57
1076977,898905.8,228.00
1081179,899795.3,220.65
1084480,900522.3,220.54
1087181,901186.2,225.00
1091382,901987.5,217.61
1097084,903147.5,214.41
1101886,904182.3,214.58
1104287,904772.6,216.90
1104587,904816.9,216.27
1130995,910489.5,215.61
1135497,911436.3,215.24
1156504,915970.3,215.39
1158604,916486.7,216.12
1163406,917472.8,215.56
1167607,918377.6,215.55
1170308,918976.1,215.72
1197617,924696.0,214.36
1200919,925430.3,214.56
1203920,926115.4,214.87
1206921,926842.2,215.48
1210522,927593.4,215.30
1211122,927724.2,215.31
1215023,928537.0,215.12
1218024,929243.5,215.54

Just working out the average visually it's about 217Kpt/s averaged for each core.
That's an approx total average of 868 Kpt/s for all 4 cores

Next I'll try 1 client auto threading.

iNSaNiTi
Did you bench v444d?
[TA]Assimilator1
2008-03-02 13:28:09
Hmm, 1 client on auto threading is 'only' using about 94-97% of CPU power most of the time .
[TA]Assimilator1
2008-03-03 18:32:53
Opps ,forgot to say rig specs (for above & below)
Intel C2Q Q6600 @ 3GHz ,RAM @ 394MHz 4-4-4-15, 2T.

Well, it seems Stephen has done a good job of improving multi-threading on DPAD, a single client is now faster than 4 single threaded clients, see results below (a turn around from v443x).

Single client ,auto threading ,approx 28 hr run.

1223131,2248826.2,1005.68
1223733,2249596.0,1142.37
1229150,2255332.2,1074.15
1229752,2256007.5,1078.14
1230354,2256632.5,1075.08
1231256,2257408.7,1052.81
1232159,2258438.8,1061.08
1233363,2259585.0,1048.98
1234266,2260485.1,1044.99
1234567,2261103.7,1070.25
1235770,2262236.3,1058.50
1236372,2262999.2,1067.59
1237275,2263908.1,1063.86
1237877,2264545.6,1063.68
1239081,2265634.9,1052.14
1239983,2266768.7,1062.68
1240585,2267383.8,1061.32
1241789,2268528.2,1054.41
1242391,2269245.1,1058.55
1242993,2270037.0,1066.12
1243895,2270936.1,1063.15
1244497,2271673.9,1067.61
1245099,2272316.8,1067.63
1246002,2273245.5,1066.13
1246905,2274147.2,1063.64
1247506,2274822.5,1065.05
1248710,2275964.2,1059.69
1249613,2276875.3,1058.00
1250215,2277604.6,1061.35
1250817,2278342.9,1064.87
1251719,2279409.6,1068.47
1252923,2280390.5,1058.43
1253224,2280906.4,1064.86
1254127,2281804.4,1062.86
1254729,2282523.0,1065.31
1255030,2282851.2,1065.54
1255932,2283607.7,1059.39
1256233,2283964.5,1060.52
1257136,2285112.9,1066.04
1257738,2285480.8,1058.26
1258340,2286200.7,1060.58
1259243,2287316.7,1064.90
1260145,2288292.1,1065.27
1261048,2289104.9,1061.41
1261650,2289844.6,1063.98
1262854,2290973.0,1060.21
1267368,2295494.5,1054.32
1268271,2296484.2,1055.14
1268872,2297208.1,1057.06
1269173,2297570.4,1058.01
1269775,2298297.7,1059.92
1270076,2298557.8,1058.69
1271280,2299736.5,1056.72
1271882,2300457.3,1058.44
1272784,2301233.8,1054.88
1273386,2302006.0,1057.58
1273988,2302657.3,1057.87
1274891,2303576.0,1057.17
1276095,2304729.2,1054.95
1276696,2305492.6,1057.32
1280007,2309211.4,1061.13
1280308,2309627.3,1062.80
1280909,2310358.8,1064.37
1281812,2311079.3,1060.32
1282715,2312229.6,1063.52
1283016,2312342.4,1060.10
1283317,2312625.9,1059.51
1284220,2313677.0,1061.05
1284828,2314406.9,1062.40
1285731,2315367.0,1062.41
1286634,2316147.0,1059.62
1287236,2317046.7,1063.67
1287838,2317549.2,1061.56
1288138,2317644.6,1058.14
1289041,2318567.3,1057.65
1289643,2319473.9,1061.68
1290546,2320214.1,1058.47
1291148,2321011.0,1060.80
1296565,2326749.9,1060.70
1297467,2327662.1,1060.10
1298370,2328637.3,1060.33
1298972,2329257.9,1060.11
1299574,2329982.9,1061.23
1300176,2330344.9,1057.67
1300778,2331253.5,1061.15
1301981,2332273.2,1057.90
1302884,2333171.4,1057.20
1303486,2333942.7,1058.86
1304389,2334934.8,1059.31
1304991,2335632.1,1060.03
1306194,2336777.6,1058.47
1309504,2340482.1,1060.78
1310106,2341150.5,1061.12
1311009,2341941.8,1059.24
1314319,2345790.3,1062.97
1315222,2346699.2,1062.42
1318532,2350420.1,1064.55
1319134,2350926.8,1063.16
1320338,2352090.9,1061.98
1320940,2352754.7,1062.23
1321541,2353525.7,1063.56
1322143,2354299.7,1064.90
1323347,2355496.0,1064.05
1324250,2356501.4,1064.49
1324852,2357196.9,1065.03
1325754,2358019.5,1063.68

If I've used excel right that's an average of 1061.51 ,that's nearly 25% faster than the previous setup!

Next to try it 2 clients with 2 threads each.

[TA]Assimilator1
2008-03-04 18:32:16
Same Q6600 @3GHz etc.

2 clients ,2 thread setting.Approx 12hr run.

Client 1
1329634,2360495.0,512.93
1330235,2360782.1,502.99
1331736,2361510.5,495.56
1332937,2362257.0,527.02
1335041,2363303.2,518.00
1336545,2364077.2,517.37
1338952,2365227.0,508.56
1340155,2365961.7,518.77
1347071,2369676.5,525.46
1348574,2370418.7,523.14
1360008,2376156.4,515.50
1360610,2376548.4,518.01
1361813,2377120.5,516.49
1363017,2377858.0,519.80
1363318,2378026.7,520.15
1364823,2378760.4,518.82

Average 516.16

Client 2
1339867,2131315.6,502.64
1341071,2131958.8,506.39
1343477,2133097.5,500.06
1345281,2134023.9,501.74
1347385,2135043.5,499.57
1348287,2135680.7,510.27
1350392,2136605.0,502.59
1352198,2137503.1,502.16
1352799,2137873.6,505.27
1354002,2138616.7,511.11
1354303,2138644.0,505.72
1356109,2139543.3,505.17
1358215,2140640.3,506.37
1358516,2140799.9,506.63
1360021,2141542.2,505.95
1361826,2142446.1,505.64
1363331,2143226.1,506.23
1364535,2143929.8,509.03
1366039,2144716.6,509.63
1366942,2145255.4,511.81
1369349,2146402.6,509.60
1370553,2147093.0,511.54

Average 506.14
Total average 1022.3

Single client rules! 
[OCAU] badger
2008-03-05 03:44:01
I just got a new machine, e8400 core2 duo, I have it running at 9x412 (3.7Ghz) at the moment, and with a single client with threads set to '2' I'm getting 570kpts/s. Have noticed that it starts out faster (at up to 591 kpts/s) and then slows down slightly over time, I'm not sure why, perhaps it is time reading the results.dat file which is increasing in size?  I'm running it on scientific linux v 5.1 with wine, it sits on 95% cpu with threads set to '2', but only at about 75% if threads are set to 'auto'.
[OCAU] badger
2008-03-05 03:49:27
Just as a comparison, my P4 3.2Ghz does 227kpts/s with version 444d (under winXP), which isn't a lot slower than half of my e8400, even though calculating 1 million decimal places of Pi with the P4 takes 38 sec, and only 12 sec with 1 core of the e8400...
I'll do some numbers with my dual opteron 2216 overnight.
Stephen Brooks
2008-03-06 01:01:15
--[it sits on 95% cpu with threads set to '2', but only at about 75% if threads are set to 'auto'.]--

EH?  I can't help thinking this might be a bug in Wine.  If you get the commandline version up when you use "auto", how many "logical CPUs" does it say it detects?  In theory auto should be the same as 2 if that is the number of cores Muon1 has detected.
[OCAU] badger
2008-03-06 02:21:58
Hmm, turns out there were some problems, the results.txt file was write protected, so there was some delay while it complained about not being able to write to it.  After running overnight again the e8400 is going faster, but the rate is still dropping off quickly.

I used muon1bench with 120 sec interval.  I always wait until there is no queue or autosave file, then kill muon1, then start it and muon1bench a second or so later.  Thus the maximum speed listed should be below the true maximum (and will be the maximum if there is no delay between the result being saved and muon1bench checking the file.  I also usually rename the results.dat file to start fresh, but I didn't in this last case and still saw the reduction in speed over time.  I'm not seeing this reduction in speeds from any other benchcsv.log posted above:
(Latest Example)
Uptime (secs),Mpts in file,Estimate kpts/sec
841,57115.7,0.00
1201,57364.6,691.39
4681,59491.4,618.67
5281,59823.6,609.89
6001,60302.5,617.60
6361,60486.1,610.58
6961,60895.8,617.66
7441,61136.0,609.14
10921,63263.9,609.94
11161,63433.9,612.23
11881,63866.3,611.47
12961,64267.7,590.10
13201,64397.7,589.16
13441,64702.4,602.12
14041,65111.6,605.75
14521,65357.8,602.49
18241,67256.6,582.81
19081,67675.5,578.94
22441,69765.0,585.62
23041,70098.2,584.80
26641,72276.7,587.64
26761,72383.6,589.04
27241,72612.9,587.01
27961,73043.1,587.29
28921,73505.2,583.67
29161,73625.2,582.96
29281,73705.0,583.31
32761,75844.2,586.73
33121,76051.1,586.60
33841,76482.7,586.88
34201,76497.7,580.99
34921,76875.0,579.79
35641,77331.1,580.90
35881,77518.0,582.26
36481,77878.7,582.58
36841,78074.7,582.19
37441,78446.3,582.80
38281,78911.2,582.14
41641,81001.0,585.42

Looking at the individual instantaneous rates, I see the problem, most of them are quite high, but the average is brought down by infrequent very low output periods, eg in the above list:

33841 76482.7 586.88
34201 76497.7 580.99
gives a speed of 41.7 kpts/s over 6 minutes


As for the 'auto' setting: with command line it states "detecting 2 logical processors" then uses 100% of one core and 92-96% on the other (using gnome system monitor, 194% cpu according to top).  It seems to swap which core is doing 100% from time to time. 

"2" threads command line is the same

"2" threads background mode (ie wine muon1 -b & is 100 and 90-92%

"auto" threads background mode is 100% and 50-80% (163% in top) with lots and lots of swapping between cores.

Running 2 instances (background or commandline) gives 100% on each core continuously.

[TA]Assimilator1
2008-03-06 21:42:19
Damn it!  ,I've just discovered that DPAD had added a 2nd 'Download sample results file' line to the config.txt files & hence presumeably overuled the '0' I set on the original line :| ,apparently it didn't like me adding 'def 24' in the brackets (I needed that as a reminder what the default time was).

Now I don't know if any of the above quad core results are accurate! 

How much does the 'Download sample results' setting effect the benchmark? 
[OCAU] badger
2008-03-07 00:03:05
I found on some machines that couldn't get the sample results from the net for some reason (firewall maybe - I can download them manually fine), that they were slowed quite significantly as it would sit and wait for a timeout every single time it started a new sim, adding ~20s to each sim (which take 3-15 minutes), so up to 10% slower...

As for my e8400, I ran 2 instances with 1 thread each overnight and got a score of 271 for one and 273 for the other, total of 544 kpts/s. This is a fair bit slower than running one instance with 2 threads, even though the cpu utilisation is higher with 2 threads... There wasn't the slowing down over time that I noticed with 2 threads though, it was pretty constant - this is strange.

I also realised that the benchmark I mentioned for the P4 3.2Ghz earlier was probably with a previous version of muon1 (it's timestamped earlier than I installed v444d), so I'll run another benchmark for comparison.
Stephen Brooks
2008-03-07 16:02:46
--[apparently it didn't like me adding 'def 24' in the brackets (I needed that as a reminder what the default time was).]--

The trouble is, if you change what is before the colon, how is Muon1 supposed to know which line is the one to look at?  (they don't have to be in any particular order).  So it generates a new default line if it can't find an exact match for one it expects to be there.

It's intentional that it checks once per simulation, so it can take advantage even if the computer is only online for a fairly short period.  However, if the computer is always going to be offline it's better to use the cfg_nonet script to stop it even trying.
[TA]Assimilator1
2008-03-07 18:42:12
Err ok, but you still haven't answered my question

Btw I didn't think comments in the brackets would effect what the program looks for.
Is their anyway to add comments in which the program can ignore?

I've just done another run with 2 clients 2 threads each .........wth?, grrrr I don't believe it!  ,one of the config files has turned download samples back on again!  ,I swear I turned them off!  ,oh well ,hell with it I'll post the client scores with the one that did stay off & just double the scores.

Client 2 Q6600 @3GHz

1606859,2152384.3,537.50
1608063,2153110.2,557.69
1609267,2153731.1,547.85
1611674,2154879.5,525.20
1622807,2160585.8,517.65
1624011,2161267.5,520.60
1626418,2162413.6,515.79
1627621,2163052.6,516.57
1628524,2163612.1,520.39
1629728,2164273.9,521.78
1641162,2169970.1,514.49
1642967,2170910.2,514.78
1643268,2170981.6,512.64
1644472,2171761.4,516.68
1645976,2172484.1,515.37
1646578,2172832.3,516.27
1647782,2173573.0,519.00
1659216,2179292.6,515.10
1659817,2179613.6,515.30
1661924,2180614.3,513.84
1662526,2180964.2,514.53
1664331,2181926.2,515.08
1665535,2182613.9,516.19
1667942,2183800.9,515.32
1668243,2183950.1,515.22
1668544,2184012.9,513.79
1669446,2184579.2,515.36

Average from excel is 520 Kpts/s, so 1040 Kpts/s for the 2 clients.
So a single client is still faster IF my earlier single client scores are accurate (retesting a single client now, 'Download sample results' is definitly off).
Stephen Brooks
2008-03-08 15:51:49
--[Err ok, but you still haven't answered my question]--

Just benchmark on an internet-connected computer, then it won't time out on every run!  My guess is that it makes very little difference except if you manage to get it stuck always timing out.

--[Is their anyway to add comments in which the program can ignore?]--

No, I originally wrote the config reader to have every line in the format <attribute>: <value>, though when looking for values I usually only read the first couple of words.  However, when setting defaults (as is necessary when people upgrade and I've added a new config option) it uses the whole string.  So make your notes in another file (or maybe below the bottom of config.txt, that might work).
[TA]Assimilator1
2008-03-08 17:58:48
Time out??  I think you're answering someone elses question
I originally asked 'How much does the 'Download sample results' setting effect the benchmark?  '

Thanks for the config info

Q6600 @3GHz
1 client ,auto threading

1675645,2663658.8,1073.81
1675946,2663908.7,1039.02
1676548,2664508.7,1029.64
1677150,2665280.4,1075.54
1678053,2666065.7,1031.45
1678655,2666822.8,1059.75
1679257,2667557.3,1077.59
1679859,2668305.1,1094.07
1681063,2669304.5,1050.08
1681965,2670437.7,1072.86
1682567,2670990.1,1062.17
1683470,2671965.8,1063.90
1684373,2672980.0,1068.98
1685276,2673743.7,1051.37
1686179,2674726.3,1054.06
1687082,2675716.6,1056.98
1688286,2676836.2,1046.40
1689189,2677809.6,1048.26
1689790,2678688.1,1063.78
1690994,2679844.3,1056.53
1691897,2680859.7,1059.93
1692499,2681264.4,1047.43
1693101,2682019.4,1053.89
1694004,2682998.0,1055.24
1694907,2683992.2,1057.20
1695810,2684785.4,1049.86
1696412,2685387.1,1048.52
1696713,2685812.2,1053.31
1697315,2686374.2,1050.24
1702130,2691278.7,1044.84
1702732,2692074.4,1050.61
1703635,2693032.4,1050.93
1704839,2694129.8,1045.52
1705741,2695291.1,1052.33
1706042,2695426.1,1046.69
1706644,2696163.0,1049.94
1707246,2696894.9,1052.93
1708450,2698083.4,1050.65
1709353,2698920.9,1047.52
1709654,2699447.2,1053.41
1710858,2700441.7,1046.02
1711761,2701616.3,1052.09
1712363,2702189.6,1050.53
1713266,2702954.6,1045.88
1714168,2704107.1,1051.04
1714770,2704677.4,1049.52
1715673,2705605.3,1049.05
1719289,2709598.9,1053.45
1720493,2710858.8,1053.28
1720794,2711136.4,1052.44
1721998,2712194.0,1048.09
1722600,2712923.5,1050.11
1723503,2714081.8,1054.35
1723804,2714186.2,1050.08
1724105,2714461.0,1049.26
1724706,2715284.1,1053.03
1725609,2716045.9,1049.38
1725910,2716585.3,1053.67
1727114,2717623.1,1049.34
1730425,2721580.3,1057.88
1731629,2722705.7,1055.32
1732531,2723711.9,1056.23
1733133,2724502.7,1058.85
1734036,2725267.9,1055.68
1734337,2725553.7,1055.15
1734939,2726159.1,1054.66
1735541,2726885.6,1056.15
1736444,2728046.1,1059.45
1737648,2729055.9,1055.29
1738250,2729849.7,1057.75
1738551,2730193.5,1058.15
1739453,2731218.9,1059.21
1740055,2731653.6,1056.15
1740657,2732335.6,1056.84
1741259,2733103.7,1058.80
1741861,2733748.3,1058.91
1745473,2737714.6,1060.89
1746075,2738437.6,1062.06
1746977,2739408.7,1062.23
1747579,2740178.8,1064.00
1748181,2740747.9,1063.04
1748482,2740850.3,1060.12
1749385,2741826.8,1060.38
1749987,2742644.8,1062.74
1750890,2743435.7,1060.55
1751793,2744445.5,1061.22
1752395,2745190.8,1062.58
1752997,2745900.7,1063.47
1753599,2746328.4,1060.80
1754201,2747072.9,1062.12
1754802,2747844.5,1063.76
1755404,2748539.1,1064.42

Average of 1055.7 Kpts/s

I guess looking at those results turning off 'download samples' doesn't make much difference.
Stephen Brooks
2008-03-08 19:37:25
I don't think it makes much difference.  Ideally you should leave it turned on and benchmark with the computer connected to the internet.  Badger said something complicated about if you have the computer disconnected from the internet then Muon1 will spend 20-30 seconds testing for timeout to see if it can contact the website and download samplefiles if they are not fresh enough, but this is not a "typical" situation for benchmarking.

Samplefiles will give you longer, higher-yielding simulations, too, which is as it should be.
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io RSS feed

Site has had 25163659 accesses.