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
Stephen Brooks
2009-03-10 17:29:45
Oh, over-*volting* may have detrimental consequences due to breakdown.  But overclocking at stock voltage with an enhanced cooler really should not.
[TA]Assimilator1
2009-03-28 10:48:20
K'Tetch

"Calculation errors are sought out & eliminated by extensive stress testing, this is fairly basic overclocking knowledge btw"
>>>>Yeah, I know.  When I was still active in the field, there was a project that asked for people with stable Overclocks to test their systems.  It used a very expensive chip analyser which took the chip, and emulated the motherboard, giving pre-programed inputs and seeing the output.  Very nice, and it found that in maybe 40% of the 'stable, extensively stress tested' cases, the chips were producing errors that went unspotted, mainly because they were areas that didn't get used often, or had redundency, or were intermittant.<<<<<

That as it maybe, but we don't know how well they tested (length of testing & choice of programs) their systems in the 1st place.

Also I refer back to the point that most (if not all?) DC projects have error checking, if no errors or problems show up in any programs then it doesn't really matter if some unused part of the chip is unstable.  If at a latter point a new program shows up this unstabilty then simply downclock it a bit, however I doubt that OCCT, F@H etc etc would miss anything significant.

>>>>Of course, most overclocking is more of a placebo effect anyway, even a 10% speedup is usually unnoticable unless you're specifically watching the workrate (or framerate etc)<<<<

Which everyone here is because we're running DPAD lol
But I know what you mean, those who just game won't notice much difference unless it's a large boost or the machine is only just fast enough for xxx game.

I play games occasionally, old systems would be useless for them.  Not to mention they'd be dog slow at DPAD etc.
I do have a Cel 366 @550 that I keep for some old games, it gets used once in a blue moon.

What spec is your QL-62?

Btw Intel wouldn't sponsor overclocking events if it meant that they got a bad rep from chips burning out (warrantee obviosuly isn't relavant).
Why can't you just face facts that sensible overclocking (as in no excessive voltages or temps) does NOT kill CPUs significantly early!

>>>and the materials used to make CPUs have radically changed in the last 10 years<<<
Actually they have , interconnects used to be made from Aluminium back then, just under 10yrs ago AMD & then Intel switched to copper.  Not to mention more recent changes like the new materials Intel use for their transistor gates (I forget the chemical now).

Regardless how stable our o/ced systems are, 1 thing remains true 'sensible' overclocking does not kill CPUs radically early*. I've been overclocking CPUs for nearly 11yrs now, how long did you?

*excluding really badly made cips like the old S7 Cyrix chips which I latter found out they were so on the edge that they would fairly commonly just die anyway without overclocking.
[TA]Assimilator1
2009-04-10 10:38:54
Hafnium is the element I was thinking of.
[TA]Assimilator1
2009-04-25 11:31:36
Back to topic with a couple of benchmarks from an AMD Phenom II rig.

Win XP x64
Phenom II 940(?) BE 3 GHz (default speed), NB freq.  1806 MHz, FSB 200 MHz, 6GB RAM DDR2 800 5-5-5-15.

60104,75963.2,0.00
61304,77274.8,1092.91
62504,78813.1,1187.35
63704,80454.1,1247.35
64304,81225.0,1252.69
65504,82857.3,1276.57
66105,83761.3,1299.56
67305,85392.5,1309.51
68805,87071.0,1276.64
69705,88360.0,1291.22
70905,89983.3,1298.04
71205,90211.0,1283.47
72405,91740.0,1282.56
73005,92602.7,1289.77
74205,94200.6,1293.32
74505,94226.1,1268.14
75705,95959.5,1281.70
76905,97578.8,1286.53
77205,97986.6,1287.81
78406,99563.3,1289.51
79306,100409.8,1273.15
80506,102038.9,1278.11

I don't understand why the uptime was at ~17hrs before it's 1st estimate
Anyway ignoring the 1st 3 lines the average is 1282.4 Mpts/s, that already just beats my Q6600 @3.34 GHz!

Same rig overclocked, Phenom II @3.48 GHz, NB 1680 MHz, FSB 240 MHz, RAM 800 MHz 5-5-5-18, (NB underclocked as I hadn't finished tweaking it, so this score just for comparision to the next).

1490,149658.3,0.00
2390,151266.7,1786.98
3290,152530.9,1595.77
3590,152991.4,1587.08
4190,153917.0,1577.18
5390,155531.0,1505.71
6290,157160.2,1562.78
7190,158358.2,1526.19
8390,160013.6,1500.66
9290,161660.1,1538.58
10490,163282.3,1513.67
10790,163844.6,1525.30
11990,165473.8,1506.13
12290,165955.0,1508.84
13491,167583.7,1493.68
14391,169212.9,1515.75
15291,170358.0,1499.87
16191,171983.8,1518.63
17091,173104.1,1502.83
17991,174732.5,1519.54
19191,176376.9,1509.42
20391,177876.6,1492.92
21291,179505.8,1507.34
22491,181140.6,1499.05
22791,181723.3,1505.29
23991,183347.2,1497.18
24291,183909.5,1502.14
25491,185564.2,1495.97
25791,186143.9,1501.36
26391,187107.8,1503.89
27592,188739.9,1497.27
28792,190426.0,1493.21
29392,191536.2,1500.89
30592,193190.9,1495.86
31492,194910.8,1508.31
32692,196539.4,1502.49
33892,198183.5,1497.58
34792,199812.7,1506.03
35992,201439.4,1500.79
36892,203068.6,1508.66
37192,203330.1,1503.30
38392,204876.5,1496.32
39292,206491.3,1503.41

Again ignoring the 1st 3 lines the average is 1510.8 Mpts/s

Same rig again this time with the NB o/ced.
Phenom II @3.48 GHz, NB 2160 MHz, FSB 240 MHz, RAM 800 MHz 5-5-5-18

88211,209724.2,0.00
89411,211353.5,1357.64
90311,212982.7,1551.54
91211,214466.2,1580.54
92411,216095.4,1516.83
93311,217724.5,1568.56
93611,217856.3,1505.83
94511,219514.5,1553.90
95711,221143.7,1522.48
96911,222774.3,1499.90
97812,224417.4,1530.43
99012,226006.5,1507.51
99912,227643.1,1531.42
101112,229260.2,1514.31
102012,230914.1,1535.38
103212,232511.7,1519.05
104112,234132.9,1535.02
105312,235934.0,1532.62
106512,237563.3,1521.15
107412,239192.5,1534.69
107712,239208.7,1511.91
108312,240093.3,1510.79
109212,241679.8,1521.58
110412,243310.5,1512.78
111012,244442.6,1522.62
112213,246071.7,1514.37
113113,247691.7,1524.69
114313,249310.8,1516.62
114913,250566.3,1529.55
115513,251162.2,1517.76
116413,252819.9,1528.10
117613,254432.2,1520.56
118813,256085.9,1514.97
119713,257702.7,1523.01
120913,259372.2,1518.17
121213,259870.0,1519.45
122413,261525.3,1514.54
123313,263131.0,1521.45
124514,264752.9,1515.83
125414,266369.5,1522.61
126614,268184.5,1522.29
127814,269764.6,1516.06
128714,271393.7,1522.59
129914,273018.7,1517.74
130814,274481.8,1520.02
132014,276109.4,1515.53
132914,277738.7,1521.46

Ignoring the 1st 2 lines the average is 1524 Mpts/s, only a slight increase, 1% faster . Assuming that isn't just natural variance.

(Btw that last RAM timing I increased from 15 to 18 as I discovered that SPD was setting it too low!)

Universal Creations

The Phenom II you tested had a higher score with a lower clock speed, what speed was its RAM?

Stephen
You didn't answer me earlier, what would you expect the natural score variation in muon1bench to be?
[TA]Assimilator1
2009-04-25 11:33:17
Re-confirms that in DPAD at least Phenom IIs are faster than C2Qs.
Universal Creations
2009-04-26 18:43:12
The difference is the OS.  Under 64-bit DPAD runs significantly slower than under 32-bit.  That's the reason I still use WinXP 32-bit at the moment and not Windows 7 64-bit (which runs beautifully, but not for DPAD).
runesk
2009-05-06 08:28:42
Xeon X5500, 2.67GHz, Windows 2003 SP2 64-bit

Threads (number or 'auto'): 8

Uptime (secs),Mpts in file,Estimate kpts/sec
5427,1991.8,0.00
6027,3290.1,2163.89
7827,7003.3,2088.18
9627,9995.7,1905.74
12927,17786.0,2105.95
16827,26285.9,2131.11
17127,26584.2,2101.97
17427,27190.1,2099.91
21627,36339.6,2120.29
22527,38210.1,2118.08
23127,39547.0,2121.81
23727,40809.5,2121.24
24327,42443.2,2140.34
24927,43696.4,2138.75
25227,43870.0,2115.11
25527,44999.9,2139.76
26427,46871.0,2137.16
27327,48346.3,2116.70
27927,50110.9,2138.68
28527,51517.5,2144.02
29427,53216.9,2134.43
30327,55044.9,2130.70
34527,64424.9,2145.52
38727,73916.5,2159.95
39027,74012.9,2143.54
39627,75901.6,2161.16
40226,76575.6,2143.27
40526,77349.7,2147.00
41126,78886.9,2153.98
42026,80690.1,2150.28
42626,82168.4,2155.34
43526,84032.4,2153.35
44126,85288.8,2152.43
45026,87186.0,2151.42
45626,88299.2,2147.00
46226,90085.0,2159.20
46826,91453.0,2160.95
47126,91594.7,2148.80
51326,101087.0,2158.99
52226,103006.0,2158.48
53126,104938.3,2158.26
54026,106820.3,2157.02
54326,107611.2,2159.96
54926,109022.5,2162.29
55826,110939.6,2161.72
56726,112810.2,2160.26
57626,114671.9,2158.68
58526,116584.3,2158.10
59426,118468.4,2157.03
63626,127960.7,2164.47
67826,137453.0,2170.91
68726,139243.1,2168.32
69626,141149.8,2167.62
[TA]Assimilator1
2009-05-09 09:43:41
Nice , what FSB & RAM speed is that running on?

Btw I seem to remember Stephen saying that with 8 threads a single instance isn't very efficient, have you tried 2 clients with 4 threads each?
Stephen Brooks
2009-08-03 15:11:32
Sorry, seems like I stopped posting these graphs and had an argument about overclocking instead.  Here's it is updated - note the scale change and the fact the last graph's #1 (1651.54) is now #5.



Can runesk confirm for me that his X5450 was a two-socket quad core whereas the "X5500" (did you mean X5550?) just above it was only one socket?  The change in efficiencies seems a bit extreme between these two.
Stephen Brooks
2009-08-04 20:45:28
Hmm that writing's got way too small now, I'll try and tweak the paper size for the next graph (maybe go portrait instead of landscape).
tomaz
2009-08-05 13:01:04
perhaps runesk's x5550 is dual socket ?  that would yield 101 pts/mclock/core.
by the way, will intel make dual i7s ? 
Stephen Brooks
2009-08-05 16:42:44
I get the impression the X55n0 series *are* the dual-socket i7s, for now, since they're the only ones with enough QPI links to go a multi-socket motherboard.  Though people on the internet appear to be asking whether they'll make a "extreme" enthusiasts board (like Skulltrail was for Core 2) that doesn't require buying Xeons.
[TN]marvik
2009-08-18 09:39:32
I firmly believe that runesk's benchmark (from 2009-05-06 where he previously stated X5500) was from a dual socket configuration Intel Xeon X5550 @ 2.67GHz, running with 24GB RAM.  We are colleagues, and in fact, I was the one that started the benchmark.  runesk posted the result
Stephen Brooks
2009-08-19 14:58:12
OK, fixed that.  (You've got the name on the chart for that one because runesk otherwise had his on twice )
Stephen Brooks
2009-08-19 15:00:31
I still don't quite get why the dual-socket Xeons aren't beating the single socket overclocked i7s above them.  Is the memory slower with FBDs?  Or is hyperthreading unavailable?
tomaz
2009-08-27 07:34:56
I think sole HT is not the reason.  My i7 with HT off got 127 pts/mclock/core in comparasion to ~100 of other CPUs.  I read it a year ago so I can't recall exactly but point is, that they incorporated new materials for i7 so transistors switch faster-that is maybe 20 % difference ?
But of course, HT works perfect on i7, busting it to 180+ pts/mclock/core !
[XS]riptide
2009-08-29 02:20:45
"I still don't quite get why the dual-socket Xeons aren't beating the single socket overclocked i7s above them.  Is the memory slower with FBDs?  Or is hyperthreading unavailable?"

Stephen.... DDR2 is better than FBDimms in terms of latency.  DDR3 is better again.  And the Xeons you speak about are not nehalem architecture.  They are hapertowns, not Gainstown.
[XS]riptide
2009-08-29 02:21:22
That should spell Gainestown.
Stephen Brooks
2009-08-29 16:52:49
I thought there was one of each - an X5450 and an X5550. (Though it's a bit hard to read without a magnifier)
[TA]Assimilator1
2009-09-06 13:22:23
Lol, yea that graph writing is way too small , thanks for the graph anyway .
Maybe you need a seperate page for the graph?

Oh btw & you've included results from previous versions of DPAD again .
Stephen Brooks
2009-09-09 12:44:33
I haven't recorded the version number for the different benchmark results (and some people didn't give it), so I treat them as the same.
RGtx
2009-09-10 01:17:01
What, almost certainly, has a greater bearing, than Muon1 version No., is OS.  It has been observed that there is a large disparity of Muon1 benchmarks between 32 and 64-bit Windows ( http://stephenbrooks.org/forum/?thread=1353&bork=qazcbvgwhn ).
Stephen Brooks
2009-09-10 15:56:07
Yes, it's worrying the loss in performance is so bad going to 64-bit because it implies a lot of other 32-bit applications will lose performance using a 64-bit operating system.  I wonder whether this is a processor architecture thing (AMD vs. Intel) or something else?
[OcUK]diogenese
2009-09-12 21:22:30
I've been running muon on my samsung NC10 with an atom N270 cpu.  it's hardly worth the effort!

Uptime (secs),Mpts in file,Estimate kpts/sec
111461,43572.8,103.01
112061,43619.6,102.84
129462,45448.6,103.21
133062,45767.9,102.74
133362,45792.1,102.68
143562,46868.3,102.92
Stephen Brooks
2009-09-13 22:50:50
You've got a Samsung NC10 too, eh?  That was my 2008 xmas present to myself.  And the score from my one is on the chart above - 107.27. It's fine for anything that doesn't require bulk calculation though!
[OcUK]diogenese
2009-09-17 23:40:29
Q6600 core2 quad at 3.15 Ghz, 9x350 fsb
memory 4-5-5-12 at 350
Win XP pro (32bit)

Uptime (secs),Mpts in file,Estimate kpts/sec
556389,3699815.8,0.00
564200,3709023.4,1178.75
566003,3710904.5,1153.41
567505,3712603.2,1150.37
568406,3713628.3,1149.41
569307,3714657.9,1148.94
570208,3715680.5,1148.02
571110,3716556.3,1137.19
572012,3717194.2,1112.37
572913,3718046.9,1103.30
573814,3719086.9,1105.93
574715,3720123.9,1108.14
576218,3721970.4,1117.30
577419,3723091.0,1106.76
577719,3723578.6,1114.04
579222,3725418.8,1121.34
580123,3726449.6,1122.20
581325,3727481.9,1109.51
581925,3728390.8,1119.00
583127,3729534.3,1111.49
584629,3731364.4,1117.17
585530,3732337.4,1116.02
587032,3733881.9,1111.72
587632,3734623.4,1114.08
588533,3735651.7,1114.84
590035,3737312.7,1114.44
590937,3738342.5,1115.18
591838,3739378.5,1116.06
592739,3740338.7,1114.81
593640,3741374.9,1115.65
594841,3742730.3,1116.04
596644,3744576.7,1111.93
598146,3746414.6,1115.95
599348,3747885.7,1118.98
600249,3748856.2,1118.12
600850,3749543.9,1118.48
602652,3751399.2,1115.00
[OcUK]diogenese
2009-09-17 23:43:55
Yes I do have a dinky samsung, I use it when away from my main computer to get my stats fix Pay as you go mobile broadband by vodafone is pretty cheap if you don't go downloading videos!
[TA]Assimilator1
2010-01-03 14:17:46
Regarding the graph, looking at a couple of the higher scoring dual cores I found that iNSaNiTi's Opteron score was from v443x (he said so in his last post 2yrs ago) & he never did post v444d scores unfortunately.  Seeing as v443 was faster his score should be removed from this graph, or at least the next.

Also OCAU Badgers C2D @3.7GHz score is below par, seems in his last posts about it that he was having some sort of problem of a diminishing score, so that score isn't representive either.

AND [XS]riptide's score on his E6600 @3.6GHz is mis-representive too because he ran it alongside SoB the wally! 

So basically the top 3 dual core scores are wrong, which leaves my old E6420, lol.
Come on lads, sort it out! 

Badger
Did you ever get to the bottom of that problem?

iNSaNiTi
You still about?
Btw I never said, awesome o/c on that Opty 170! 
Have you still got it to run a v444d benchmark?


Anyhow I'm about to post the fastest dual core score (for v444d anyway).

This is for my 2nd rig a Pentium E5200 @3.5 GHz, 333MHz FSB, RAM @400 MHz, 4-4-4-15, P45 chipset. 
(Inccidently these Pentiums have a Wolfdale core with 2MB cache, so really they should be called C2Ds )

Uptime (secs),Mpts in file,Estimate kpts/sec
1894,1195871.5,0.00
2195,1195985.2,378.44
2796,1196525.3,725.36
3697,1197128.1,697.07
3997,1197298.6,678.56
4598,1197882.9,743.85
8804,1200868.9,723.18
9405,1201175.5,706.15
10307,1201784.7,702.91
10607,1202074.2,711.89
11509,1202669.7,707.09
15715,1205655.7,707.95
16316,1206235.8,718.67
20522,1209224.0,716.81
21423,1209813.8,713.93
25630,1212813.0,713.77
26230,1213392.1,719.94
27132,1213982.9,717.64
27432,1214202.1,717.78
31638,1217179.4,716.37
32540,1217797.3,715.46
33441,1218458.4,715.98
34042,1218942.0,717.64
38248,1221966.5,717.80
39150,1222650.2,718.79
40051,1223250.0,717.53
40952,1223853.9,716.43
41253,1224078.1,716.66
42154,1224664.2,715.17
42755,1225272.6,719.54
46961,1228289.8,719.34

That's an uptime of 13hrs+. Run on XP 32bit.
Ignoring the 1st 2 lines obviously, the average is 691.9753333, that's 692 for the graph

Just realised I haven't tweaked the tRD yet, guess that'll be next score at some point.....
[TA]Assimilator1
2010-01-03 14:37:33
Stephen
I think you need to edit your 1st post so people know how to use this benchmark properly.  You want something like this:-

Run the benchmark for at least 12hrs, be sure not to run any other CPU intensive programs alongside (like games, other DC projects, video/music encoding etc).

When posting scores for the graph paste the entire benchmark log relating to that run.
Also include CPU type & speed, FSB & RAM speed, OS version (incl 32 or 64 bit) & for multi core CPUs whether DPAD is on auto threads or the number of threads.  (Auto is generally faster upto 4 cores).
Also if known, RAM timings & any chipset tweaks that have been done (like tRD or command), these aren't essential but will explain any speed boosts over default.

***********************************************************

I think that covers it, unless anyone else can see something I've missed?
Btw, really this should of been sorted out years ago.......
[TA]Assimilator1
2010-01-03 14:39:58
I know what I forgot to ask for, chipset type.  (Pretty much all hardware info can be found out using CPU-Z btw)
[TA]Assimilator1
2010-01-03 14:40:59
Grrr, lack of editing .....

My Pentium E5200 is running on a P43 chipset not a P45 as I said earlier.
[TA]Assimilator1
2010-01-03 14:44:39
Aww crap!  ,bloody excel, the score for my E5200 is wrong (I thought it looked wrong :$ )

The average is 714.8 not 692!
[TA]Assimilator1
2010-01-03 14:55:35
Interesting comparing my E5200 to my old E6420, it's only clocked ~10% faster (with a slower FSB & 1/2 the cache) yet is ~24% faster in DPAD . And the E5200 uses less power, 132w vs ~180w at the wall.

I guess the CPU architectural improvements have had quite an impact, although the E6420 was running on an Nvidia 650 chipset vs an Intel P43 which will contribute I guess.
Stephen Brooks
2010-01-05 18:03:21
[TA]Assimilator1, seeing as I don't have any more information than you do (just what's on this thread), can you make a list of the old-version scores that I should be removing next time I make a graph?  (we don't have a shortage of bars on it )

In case it helps, v4.44d was published 2008-Jan-09, so anything before that can't be from that version.  I keep meaning to find out why exactly 4.44d was slower than 4.43d but I'm still writing up and correcting my thesis for a few weeks yet.

I can't edit posts either, I wrote this bulletin board and never added an edit function!
RGtx
2010-01-06 20:52:23
I don't recall 4.44d being slower than 4.43d, the latter being dropped as it gave differing results to 4.44d for a given trial, screwing up the optimizations of those using sample files.
Now, 4.45 was much slower than either 4.43 of 4.44, and was subsequently dropped after a couple of weeks.
[TA]Assimilator1
2010-01-07 07:28:19
v4.44d is definatly slower than v4.43x in the few benchmarks I've seen, my old Sempron @2.5GHz S754, did ~260 Kpt/s on 4.43, ~230 Kpt/s on 4.44d. I think I have more benchmarks which show the same & I think their are some within this thread too.  I never saw 4.45, sure you don't mean 4.44e?

Stephen
Yea no probs, I'll give it a shot, however it'll take me some weeks as I don't have much spare time (unless work's closed today ).
And yes the date helps , you should put the release dates in the version history.

I thought you were the only 1 who could edit posts & I'm sure I've seen you do it before?
(PS clever not to have an edit function lol)
RGtx
2010-01-07 11:20:39
Your correct, I did mean 4.44e not 4.45. Perhaps what also needed to be taken into account was the trials actually being run, for example, those runs with a resulting lower muon% will on a multi-core CPU running a single instance be running on a single core towards the end, for a greater proportion of time (Not to mention, the greater number of writes to HD).  So a benchmark run using a recently introduced optimization will give a lower benchmark score.
Perhaps, what also needs to be investigated, is the reason why performance is so poor on win64 variants.
[TA]Assimilator1
2010-01-07 17:48:32
Damn, DP!  :$

Interesting, I didn't know that a multi-core rig could end up running on a single core on some sim's, how did you come across this?
That being the case then it makes the benchmark considerabley harder to use properly, e.g. at what point is it too early to run a benchmark on a new optimisation?
[TA]Assimilator1
2010-01-09 11:57:14
RGtx

I've just been looking at the benchlog file from my 2nd PC (which was the Sempron 3100 & is now a Pentium dual core).

It looks like the score of 258 Kpt/s might of been from v4.44c & not v4.43 as I was thinking above , I'll have to trawl through the thread to find my original post for that benchmark.

Stephen
I've begun to complie a list
RGtx
2010-01-09 14:03:45
Last night, I used a quick & dirty .bat to time an identical queue file under v4.43d, v4.44d, & v4.44e. The results look as though they may need more investigation as v4.44d appears faster than v4.43d, and the performance of v4.44e looks more lamentable than previously reported:

C:\Users\RGtx\Desktop\v443d>echo.  | Time
The current time is: 20:31:37.59
Enter the new time:

C:\Users\RGtx\Desktop\v443d>muon1 -b -q

Muon1 started: configuring...
C:\Users\RGtx\Desktop\v443d>echo.  | Time
The current time is: 21:13:06.83
Enter the new time:

C:\Users\RGtx\Desktop\v443d>Pause
Press any key to continue . . .

tantalumrodr=000;tantalumrodz=588;d10l=000;d11l=000;d12l=000;d13l=000;d14l=000;d15l=000;d16l=000;d17l=002;d18l=003;d19l=000;d1l=000;d20l=009;d21l=014;d22l=012;d23l=002;d2l=618;d3l=035;d4l=011;d5l=008;d6l=000;d7l=003;d8l=003;d9l=004;pd1=000;pd10=000;pd11=000;pd12=000;pd13=000;pd14=000;pd15=000;pd16=000;pd17=000;pd18=000;pd19=000;pd2=000;pd20=000;pd21=000;pd22=000;pd23=000;pd24=000;pd25=000;pd26=000;pd27=000;pd28=000;pd29=000;pd3=000;pd30=000;pd4=000;pd5=000;pd6=000;pd7=000;pd8=000;pd9=000;phaserotcells=491;prf10p=179;prf10v=974;prf11p=182;prf11v=977;prf12p=186;prf12v=992;prf13p=185;prf13v=977;prf14p=186;prf14v=977;prf15p=186;prf15v=973;prf16p=186;prf16v=970;prf17p=188;prf17v=989;prf18p=187;prf18v=990;prf19p=191;prf19v=974;prf1p=182;prf1v=978;prf20p=188;prf20v=973;prf21p=188;prf21v=992;prf22p=187;prf22v=971;prf23p=186;prf23v=976;prf24p=190;prf24v=976;prf25p=187;prf25v=972;prf26p=186;prf26v=976;prf27p=189;prf27v=992;prf28p=190;prf28v=988;prf29p=191;prf29v=978;prf2p=179;prf2v=979;prf30p=190;prf30v=977;prf3p=182;prf3v=984;prf4p=181;prf4v=991;prf5p=182;prf5v=975;prf6p=078;prf6v=990;prf7p=181;prf7v=977;prf8p=180;prf8v=978;prf9p=181;prf9v=988;ps10f=116;ps10l=975;ps11f=117;ps11l=993;ps12f=117;ps12l=994;ps13f=119;ps13l=991;ps14f=118;ps14l=981;ps15f=114;ps15l=979;ps16f=112;ps16l=982;ps17f=114;ps17l=983;ps18f=115;ps18l=979;ps19f=116;ps19l=976;ps1f=121;ps1l=969;ps20f=115;ps20l=971;ps21f=115;ps21l=974;ps22f=117;ps22l=974;ps23f=113;ps23l=978;ps24f=109;ps24l=976;ps25f=106;ps25l=970;ps26f=101;ps26l=971;ps27f=100;ps27l=990;ps28f=109;ps28l=990;ps29f=120;ps29l=969;ps2f=121;ps2l=969;ps30f=131;ps30l=970;ps3f=119;ps3l=972;ps4f=114;ps4l=970;ps5f=117;ps5l=990;ps6f=115;ps6l=976;ps7f=116;ps7l=977;ps8f=118;ps8l=977;ps9f=118;ps9l=975;s10f=988;s10l=987;s10r=984;s11f=985;s11l=985;s11r=976;s12f=986;s12l=986;s12r=976;s13f=986;s13l=987;s13r=973;s14f=986;s14l=987;s14r=973;s15f=985;s15l=979;s15r=985;s16f=984;s16l=971;s16r=984;s17f=981;s17l=974;s17r=973;s18f=987;s18l=976;s18r=975;s19f=978;s19l=980;s19r=973;s1f=997;s1l=971;s20f=972;s20l=988;s20r=974;s21f=959;s21l=982;s21r=990;s22f=949;s22l=944;s22r=989;s23f=951;s23l=983;s23r=965;s24f=952;s24l=991;s24r=949;s2f=986;s2l=884;s2r=932;s3f=988;s3l=791;s3r=945;s4f=989;s4l=745;s4r=964;s5f=986;s5l=870;s5r=970;s6f=000;s6l=941;s6r=987;s7f=984;s7l=968;s7r=983;s8f=981;s8l=978;s8r=983;s9f=991;s9l=981;s9r=979;#runs=5;
2.810105 (2123.8 Mpts) [v4.43d] <PhaseRotC> {396B41FAEEAF73A314C6F327}

t2-t1 = 2488.41 sec.  (853.48 KPt/s)



C:\Users\RGtx\Desktop\v444d>echo.  | Time
The current time is: 21:30:25.13
Enter the new time:

C:\Users\RGtx\Desktop\v44d>muon1 -b -q

Muon1 started: configuring...
C:\Users\RGtx\Desktop\v444d>echo.  | Time
The current time is: 22:39:35.24
Enter the new time:

C:\Users\RGtx\Desktop\v444d>Pause
Press any key to continue . . .

tantalumrodr=000;tantalumrodz=588;d10l=000;d11l=000;d12l=000;d13l=000;d14l=000;d15l=000;d16l=000;d17l=002;d18l=003;d19l=000;d1l=000;d20l=009;d21l=014;d22l=012;d23l=002;d2l=618;d3l=035;d4l=011;d5l=008;d6l=000;d7l=003;d8l=003;d9l=004;pd1=000;pd10=000;pd11=000;pd12=000;pd13=000;pd14=000;pd15=000;pd16=000;pd17=000;pd18=000;pd19=000;pd2=000;pd20=000;pd21=000;pd22=000;pd23=000;pd24=000;pd25=000;pd26=000;pd27=000;pd28=000;pd29=000;pd3=000;pd30=000;pd4=000;pd5=000;pd6=000;pd7=000;pd8=000;pd9=000;phaserotcells=491;prf10p=179;prf10v=974;prf11p=182;prf11v=977;prf12p=186;prf12v=992;prf13p=185;prf13v=977;prf14p=186;prf14v=977;prf15p=186;prf15v=973;prf16p=186;prf16v=970;prf17p=188;prf17v=989;prf18p=187;prf18v=990;prf19p=191;prf19v=974;prf1p=182;prf1v=978;prf20p=188;prf20v=973;prf21p=188;prf21v=992;prf22p=187;prf22v=971;prf23p=186;prf23v=976;prf24p=190;prf24v=976;prf25p=187;prf25v=972;prf26p=186;prf26v=976;prf27p=189;prf27v=992;prf28p=190;prf28v=988;prf29p=191;prf29v=978;prf2p=179;prf2v=979;prf30p=190;prf30v=977;prf3p=182;prf3v=984;prf4p=181;prf4v=991;prf5p=182;prf5v=975;prf6p=078;prf6v=990;prf7p=181;prf7v=977;prf8p=180;prf8v=978;prf9p=181;prf9v=988;ps10f=116;ps10l=975;ps11f=117;ps11l=993;ps12f=117;ps12l=994;ps13f=119;ps13l=991;ps14f=118;ps14l=981;ps15f=114;ps15l=979;ps16f=112;ps16l=982;ps17f=114;ps17l=983;ps18f=115;ps18l=979;ps19f=116;ps19l=976;ps1f=121;ps1l=969;ps20f=115;ps20l=971;ps21f=115;ps21l=974;ps22f=117;ps22l=974;ps23f=113;ps23l=978;ps24f=109;ps24l=976;ps25f=106;ps25l=970;ps26f=101;ps26l=971;ps27f=100;ps27l=990;ps28f=109;ps28l=990;ps29f=120;ps29l=969;ps2f=121;ps2l=969;ps30f=131;ps30l=970;ps3f=119;ps3l=972;ps4f=114;ps4l=970;ps5f=117;ps5l=990;ps6f=115;ps6l=976;ps7f=116;ps7l=977;ps8f=118;ps8l=977;ps9f=118;ps9l=975;s10f=988;s10l=987;s10r=984;s11f=985;s11l=985;s11r=976;s12f=986;s12l=986;s12r=976;s13f=986;s13l=987;s13r=973;s14f=986;s14l=987;s14r=973;s15f=985;s15l=979;s15r=985;s16f=984;s16l=971;s16r=984;s17f=981;s17l=974;s17r=973;s18f=987;s18l=976;s18r=975;s19f=978;s19l=980;s19r=973;s1f=997;s1l=971;s20f=972;s20l=988;s20r=974;s21f=959;s21l=982;s21r=990;s22f=949;s22l=944;s22r=989;s23f=951;s23l=983;s23r=965;s24f=952;s24l=991;s24r=949;s2f=986;s2l=884;s2r=932;s3f=988;s3l=791;s3r=945;s4f=989;s4l=745;s4r=964;s5f=986;s5l=870;s5r=970;s6f=000;s6l=941;s6r=987;s7f=984;s7l=968;s7r=983;s8f=981;s8l=978;s8r=983;s9f=991;s9l=981;s9r=979;#runs=5;
2.999670 (4141.0 Mpts) [v4.44d] <PhaseRotC> {D359C0AB9B9F269F0B53CAD5}

t2-t1 = 4149.87 sec.  (997.86 KPt/s)



C:\Users\RGtx\Desktop\v444e>echo.  | Time
The current time is: 22:51:04.83
Enter the new time:

C:\Users\RGtx\Desktop\v44e>muon1 -b -q

Muon1 started: configuring...
C:\Users\RGtx\Desktop\v444e>echo.  | Time
The current time is: 0:36:06.22
Enter the new time:

C:\Users\RGtx\Desktop\v444e>Pause
Press any key to continue . . .

tantalumrodr=000;tantalumrodz=588;d10l=000;d11l=000;d12l=000;d13l=000;d14l=000;d15l=000;d16l=000;d17l=002;d18l=003;d19l=000;d1l=000;d20l=009;d21l=014;d22l=012;d23l=002;d2l=618;d3l=035;d4l=011;d5l=008;d6l=000;d7l=003;d8l=003;d9l=004;pd1=000;pd10=000;pd11=000;pd12=000;pd13=000;pd14=000;pd15=000;pd16=000;pd17=000;pd18=000;pd19=000;pd2=000;pd20=000;pd21=000;pd22=000;pd23=000;pd24=000;pd25=000;pd26=000;pd27=000;pd28=000;pd29=000;pd3=000;pd30=000;pd4=000;pd5=000;pd6=000;pd7=000;pd8=000;pd9=000;phaserotcells=491;prf10p=179;prf10v=974;prf11p=182;prf11v=977;prf12p=186;prf12v=992;prf13p=185;prf13v=977;prf14p=186;prf14v=977;prf15p=186;prf15v=973;prf16p=186;prf16v=970;prf17p=188;prf17v=989;prf18p=187;prf18v=990;prf19p=191;prf19v=974;prf1p=182;prf1v=978;prf20p=188;prf20v=973;prf21p=188;prf21v=992;prf22p=187;prf22v=971;prf23p=186;prf23v=976;prf24p=190;prf24v=976;prf25p=187;prf25v=972;prf26p=186;prf26v=976;prf27p=189;prf27v=992;prf28p=190;prf28v=988;prf29p=191;prf29v=978;prf2p=179;prf2v=979;prf30p=190;prf30v=977;prf3p=182;prf3v=984;prf4p=181;prf4v=991;prf5p=182;prf5v=975;prf6p=078;prf6v=990;prf7p=181;prf7v=977;prf8p=180;prf8v=978;prf9p=181;prf9v=988;ps10f=116;ps10l=975;ps11f=117;ps11l=993;ps12f=117;ps12l=994;ps13f=119;ps13l=991;ps14f=118;ps14l=981;ps15f=114;ps15l=979;ps16f=112;ps16l=982;ps17f=114;ps17l=983;ps18f=115;ps18l=979;ps19f=116;ps19l=976;ps1f=121;ps1l=969;ps20f=115;ps20l=971;ps21f=115;ps21l=974;ps22f=117;ps22l=974;ps23f=113;ps23l=978;ps24f=109;ps24l=976;ps25f=106;ps25l=970;ps26f=101;ps26l=971;ps27f=100;ps27l=990;ps28f=109;ps28l=990;ps29f=120;ps29l=969;ps2f=121;ps2l=969;ps30f=131;ps30l=970;ps3f=119;ps3l=972;ps4f=114;ps4l=970;ps5f=117;ps5l=990;ps6f=115;ps6l=976;ps7f=116;ps7l=977;ps8f=118;ps8l=977;ps9f=118;ps9l=975;s10f=988;s10l=987;s10r=984;s11f=985;s11l=985;s11r=976;s12f=986;s12l=986;s12r=976;s13f=986;s13l=987;s13r=973;s14f=986;s14l=987;s14r=973;s15f=985;s15l=979;s15r=985;s16f=984;s16l=971;s16r=984;s17f=981;s17l=974;s17r=973;s18f=987;s18l=976;s18r=975;s19f=978;s19l=980;s19r=973;s1f=997;s1l=971;s20f=972;s20l=988;s20r=974;s21f=959;s21l=982;s21r=990;s22f=949;s22l=944;s22r=989;s23f=951;s23l=983;s23r=965;s24f=952;s24l=991;s24r=949;s2f=986;s2l=884;s2r=932;s3f=988;s3l=791;s3r=945;s4f=989;s4l=745;s4r=964;s5f=986;s5l=870;s5r=970;s6f=000;s6l=941;s6r=987;s7f=984;s7l=968;s7r=983;s8f=981;s8l=978;s8r=983;s9f=991;s9l=981;s9r=979;#runs=5;
3.012347 (2388.8 Mpts) [v4.44e] <PhaseRotC> {0AA9D85E43AB1FE3B2D926E9}

t2-t1 = 6301.17 sec.  (379.10 KPt/s)
[TA]Assimilator1
2010-01-09 14:37:49
I've found that the 260 Kpt/s score for my Sempron 3100 was from v4.43d not any v4.44 variants.

v4.44e doesn't matter anymore as it was removed after a few weeks, so don't worry about that 1 .

I don't understand how you've got better scores from 4.44d, looking back through this thread when v4.44 was released (Jan 08) a few of us found that it was 5-15% slower IIRC.

Btw, is the que file the one with the simulation in it?

Stephen
What is the normal score variance when running muon1bench?
RGtx
2010-01-09 15:46:13
Yes, the queue file does contain the simulation to be run, see Stephen's FAQ (" // Manual Seeding of Designs" for details of how to manually run simulations.  The muon1 -q switch runs the simulation in the queue file then exits.  "echo.  | Time" reads the time both pre- and post-simulation (ignoring the "Enter the New Time". Thus, permitting a fairly accurate timing of a simulation.

I should add, that in each case, there were no sample or results.dat files to be read, which would otherwise delay the start of the simulation, and "Auto-save interval" was disabled, i.e. set to 0.

What I will do next, is time a number of simulations of approx equal MPt scoring for a range of different optimizations, using v4.44d.
RGtx
2010-01-09 15:49:29
The smilies in the above post were entirely unintentional, if not appropriate.
[TA]Assimilator1
2010-01-09 15:53:55
Should be interesting, thanks for your efforts btw .

Stephen
I've finally complied the list, it's taken me hours!
Here's the list of PRE v4.44d scores to be removed from graph.

iNSaNiTi Opteron 170 - (23/01/08 v4.43d)
tomaz Athlon64 X2 4800 (11/07)
My E6420 @3.2GHz (08/07)
[XS]riptide C2Q QX6700 @3.69GHz (07/07)
tomaz Turion64 (07/07)
tomaz Athlon64 X2 5200 (06/07)
[DPC] Eclipse~Lord Alderaan 2x Xeon X2 5120 (06/07)
My E6420 @2.8GHz (06/07)
Haiya-Dragon 2x Xeon x2 5150 - (05/07)
TurtleBlue Athlon64 X2 - (05/07)
[DPC] Eclipse~Flandre 2x Opteron - (05/07)
iNSaNiTi Athlon64 3000 - (05/07, v4.43d)
waffleironhead P4 - (04/07)
rhughart C2D E6400 - (04/07)
[TA]JonB Opteron 165 - (03/07)
TurtleBlue P4 - (03/07)
bzm QX6700 - (01/07 v4.43d)
[OcUK]diogenese X2 4400 - (01/07, v4.43d)
HuubNWW X6800 - (10/06 v4.43d)
[TP]Electrolyte P4 3.2GHz - (aka [TP]Skatoony 08/06)
[TP]Electrolyte Ath64 FX-55 - (aka [TP]Skatoony 08/06)
[OCAU]Badger Athlon Barton XP2500 @2.14GHz - (07/06)
[OCAU]Badger P4 HT 3.4GHz - (07/06)
Pascal Athlon64 X2 4200 - (from graph 05/07)
[TA]Caferace C2D E4500 - (30/11/07 v4.43d)
[TP]Skatoony Athlon64 X2 4800 (05/07)
Xanathorn Athlon64 X2 4400 (06/07)
My Athlon XPM (2/06)

My Sempron 3100 score needs updating, v4.44d score is 235

Problem benchmarks which shouldn't of been included in the 1st place :-

OCAU Badgers C2D E8400 @3.7GHz - had DPAD speed probs
[XS]riptide's score on his E6600 @3.6GHz - ran with SoB!  (also 04/07 so pre v4.44d)

Also thanks to me (lol) I've noted in the thread (13/12/07) when a new client was released , v4.44 (presumeably the 1st 1 with no suffix??).

Scores dropped/missed that could be included:-

My C2Q Q6600 @3GHz, RAM@394 MHz,4-4-4-15-2T, Nvidia 650 chipset,WinXP 32bit,v4.44d - 1056 Kpt/s (see bottom of p7 for full log).

Runesk Xeon X5500 @2.67GHz (this page, just before the current graph).

That list cuts out a LOT of scores!  lol, so rather than just lose the v4.43d scores maybe they could just be put in a new seperate graph, that way older rigs could still be compared & without mixing it up with v4.44d scores.

..... I gotta go shopping right now as I've notice it's started snowing again, I'll post latter if I need to add anything!  (List is done anyway).
[TA]Assimilator1
2010-01-09 17:21:23
RGtx
Yea using the same simulations for different configs makes sense (although is your one a 'typical' one?).
That reminds me, a few of us sometime back asked Stephen if we could have some sort of 'standard' simulation purely for benchmarking purposes, now would be a good time to remind him .

Stephen
Is that possible for you to offer for download a 'typical' simulation for benchmarking purposes?  It would make the benchmark more accurate if we could all run the same one.
[TA]Assimilator1
2010-01-09 18:37:59
Oh & here's another benchmark to add from my brothers PC which is a very similiar spec to my 2nd PC (I built both).

Intel Pentium E5200 @3.7 GHz, 336 MHz FSB, 420MHz RAM 4-4-4-15, tRD 6, Intel P43 chipset, WinXP 32bit, v4.44d

Uptime (secs),Mpts in file,Estimate kpts/sec
1424,42718.7,0.00
3524,44398.7,800.01
5924,46332.7,803.13
8324,48218.2,797.04
9224,48889.7,791.17
11624,50883.8,800.51
14024,52807.2,800.69
26324,62501.7,794.51
28124,63970.7,795.97

Average 798 Kpts/s (ignoring the 1st line), the fastest dual core even over v4.43d scores!  .
Hmm, a 7.8hr run, I hope that's ok?
And yea he got a better CPU than me .
[TA]Assimilator1
2010-01-09 19:29:40
Incase I wasn't clear what I meant about the graphs,1 was for v4.44 & 1 for v4.43d.
RGtx
2010-01-09 21:55:47
After a single run, I do not think anything constructive will be achieved.

Starting from a much earlier optimization, and intending to progress to the latest, I selected a ChicaneLinacB trial from my archives and reran under v4,44d:

C:\Users\RGtx\Desktop\v444d>echo.  | Time
The current time is: 18:27:46.89
Enter the new time:

C:\Users\RGtx\Desktop\v444d>muon1 -b -q

Muon1 started: configuring...
C:\Users\RGtx\Desktop\v444d>echo.  | Time
The current time is: 21:32:48.63
Enter the new time:

C:\Users\RGtx\Desktop\v444d>Pause
Press any key to continue . . .


d10l=000;d11l=000;d12l=000;d13l=000;d14l=000;d15l=018;d16l=005;d17l=000;d18l=000;d19l=000;d1l=000;d20l=000;d21l=000;d22l=000;d23l=000;d24l=000;d2l=000;d3l=000;d4l=000;d5l=319;d6l=000;d7l=000;d8l=194;d9l=000;ld1=000;ld10=000;ld11=000;ld12=000;ld13=000;ld14=000;ld15=000;ld16=000;ld17=000;ld18=000;ld19=000;ld2=000;ld20=000;ld21=000;ld22=000;ld23=000;ld24=000;ld25=000;ld26=000;ld27=000;ld28=000;ld29=000;ld3=000;ld30=000;ld31=000;ld32=000;ld33=000;ld34=000;ld35=000;ld36=000;ld37=000;ld38=000;ld39=000;ld4=000;ld40=000;ld41=000;ld42=000;ld43=000;ld44=000;ld45=000;ld46=000;ld47=000;ld48=000;ld49=000;ld5=000;ld50=000;ld51=000;ld52=000;ld53=000;ld54=000;ld55=000;ld56=000;ld57=000;ld58=000;ld59=000;ld6=000;ld60=000;ld61=000;ld7=000;ld8=000;ld9=000;linaccells=000;ls10f=917;ls10l=892;ls11f=916;ls11l=905;ls12f=920;ls12l=897;ls13f=908;ls13l=892;ls14f=893;ls14l=921;ls15f=906;ls15l=910;ls16f=913;ls16l=909;ls17f=907;ls17l=943;ls18f=901;ls18l=902;ls19f=915;ls19l=866;ls1f=838;ls1l=888;ls20f=916;ls20l=870;ls21f=916;ls21l=921;ls22f=913;ls22l=910;ls23f=903;ls23l=928;ls24f=920;ls24l=915;ls25f=930;ls25l=902;ls26f=915;ls26l=909;ls27f=911;ls27l=900;ls28f=918;ls28l=910;ls29f=909;ls29l=896;ls2f=903;ls2l=915;ls30f=899;ls30l=902;ls31f=921;ls31l=910;ls32f=943;ls32l=984;ls33f=940;ls33l=901;ls34f=904;ls34l=931;ls35f=899;ls35l=905;ls36f=907;ls36l=910;ls37f=928;ls37l=907;ls38f=905;ls38l=908;ls39f=911;ls39l=906;ls3f=909;ls3l=925;ls40f=902;ls40l=923;ls41f=931;ls41l=904;ls42f=909;ls42l=911;ls43f=910;ls43l=909;ls44f=908;ls44l=994;ls45f=899;ls45l=998;ls46f=894;ls46l=922;ls47f=911;ls47l=911;ls48f=927;ls48l=931;ls49f=905;ls49l=931;ls4f=910;ls4l=966;ls50f=889;ls50l=899;ls51f=907;ls51l=923;ls52f=908;ls52l=905;ls53f=912;ls53l=891;ls54f=903;ls54l=905;ls55f=896;ls55l=920;ls56f=912;ls56l=931;ls57f=907;ls57l=887;ls58f=892;ls58l=947;ls59f=900;ls59l=954;ls5f=925;ls5l=897;ls60f=908;ls60l=917;ls61f=916;ls61l=947;ls6f=906;ls6l=905;ls7f=913;ls7l=939;ls8f=913;ls8l=997;ls9f=908;ls9l=898;rf10p=349;rf10v=720;rf11p=333;rf11v=686;rf12p=374;rf12v=715;rf13p=381;rf13v=997;rf14p=358;rf14v=665;rf15p=400;rf15v=738;rf16p=431;rf16v=698;rf17p=422;rf17v=645;rf18p=439;rf18v=654;rf19p=422;rf19v=680;rf1p=320;rf1v=676;rf20p=412;rf20v=682;rf21p=379;rf21v=697;rf22p=354;rf22v=995;rf23p=377;rf23v=668;rf24p=363;rf24v=750;rf25p=400;rf25v=758;rf26p=378;rf26v=662;rf27p=376;rf27v=691;rf28p=361;rf28v=777;rf29p=391;rf29v=756;rf2p=329;rf2v=999;rf30p=388;rf30v=684;rf31p=397;rf31v=739;rf32p=383;rf32v=707;rf33p=359;rf33v=999;rf34p=355;rf34v=735;rf35p=312;rf35v=722;rf36p=336;rf36v=659;rf37p=345;rf37v=718;rf38p=480;rf38v=732;rf39p=395;rf39v=674;rf3p=387;rf3v=294;rf40p=411;rf40v=000;rf41p=400;rf41v=718;rf42p=387;rf42v=682;rf43p=397;rf43v=786;rf44p=415;rf44v=999;rf45p=404;rf45v=999;rf46p=402;rf46v=694;rf47p=395;rf47v=797;rf48p=379;rf48v=755;rf49p=422;rf49v=660;rf4p=389;rf4v=714;rf50p=406;rf50v=675;rf51p=408;rf51v=595;rf52p=431;rf52v=683;rf53p=419;rf53v=733;rf54p=481;rf54v=753;rf55p=443;rf55v=724;rf56p=420;rf56v=761;rf57p=353;rf57v=773;rf58p=321;rf58v=661;rf59p=000;rf59v=669;rf5p=355;rf5v=655;rf60p=361;rf60v=690;rf61p=299;rf61v=634;rf6p=386;rf6v=999;rf7p=344;rf7v=742;rf8p=308;rf8v=996;rf9p=307;rf9v=947;s10f=900;s10l=900;s10r=909;s11f=945;s11l=917;s11r=990;s12f=897;s12l=906;s12r=897;s13f=901;s13l=893;s13r=999;s14f=898;s14l=901;s14r=906;s15f=900;s15l=909;s15r=999;s16f=911;s16l=914;s16r=903;s17f=908;s17l=898;s17r=895;s18f=900;s18l=908;s18r=926;s19f=929;s19l=910;s19r=893;s1f=997;s1l=768;s20f=918;s20l=909;s20r=917;s21f=898;s21l=898;s21r=880;s22f=935;s22l=914;s22r=925;s23f=901;s23l=906;s23r=937;s24f=773;s24l=912;s24r=521;s25f=918;s25l=885;s25r=912;s2f=971;s2l=991;s2r=913;s3f=900;s3l=688;s3r=895;s4f=907;s4l=856;s4r=929;s5f=914;s5l=985;s5r=895;s6f=912;s6l=910;s6r=900;s7f=888;s7l=907;s7r=926;s8f=890;s8l=917;s8r=891;s9f=907;s9l=911;s9r=999;tantalumrodr=000;tantalumrodz=432;#runs=5;
1.142190 (20513.2 Mpts) [v4.44d] <ChicaneLinacB> {6CDC80366A7AEE287E42E8D7}


t2-t1 = 11146.89 sec.  (1840.26 KPt/s)


I should add Q6600 2.4GHz Win 7 x64

So, it would appear that the benchmark score is also heavily dependent on which optimization is currently being run.

(Incidentally, the original file used, previously gave a much lower MPt for 5runs:

#gen=2;#runs=5;d10l=000;d11l=000;d12l=000;d13l=000;d14l=000;d15l=018;d16l=005;d17l=000;d18l=000;d19l=000;d1l=000;d20l=000;d21l=000;d22l=000;d23l=000;d24l=000;d2l=000;d3l=000;d4l=000;d5l=319;d6l=000;d7l=000;d8l=194;d9l=000;ld1=000;ld10=000;ld11=000;ld12=000;ld13=000;ld14=000;ld15=000;ld16=000;ld17=000;ld18=000;ld19=000;ld2=000;ld20=000;ld21=000;ld22=000;ld23=000;ld24=000;ld25=000;ld26=000;ld27=000;ld28=000;ld29=000;ld3=000;ld30=000;ld31=000;ld32=000;ld33=000;ld34=000;ld35=000;ld36=000;ld37=000;ld38=000;ld39=000;ld4=000;ld40=000;ld41=000;ld42=000;ld43=000;ld44=000;ld45=000;ld46=000;ld47=000;ld48=000;ld49=000;ld5=000;ld50=000;ld51=000;ld52=000;ld53=000;ld54=000;ld55=000;ld56=000;ld57=000;ld58=000;ld59=000;ld6=000;ld60=000;ld61=000;ld7=000;ld8=000;ld9=000;linaccells=000;ls10f=917;ls10l=892;ls11f=916;ls11l=905;ls12f=920;ls12l=897;ls13f=908;ls13l=892;ls14f=893;ls14l=921;ls15f=906;ls15l=910;ls16f=913;ls16l=909;ls17f=907;ls17l=943;ls18f=901;ls18l=902;ls19f=915;ls19l=866;ls1f=838;ls1l=888;ls20f=916;ls20l=870;ls21f=916;ls21l=921;ls22f=913;ls22l=910;ls23f=903;ls23l=928;ls24f=920;ls24l=915;ls25f=930;ls25l=902;ls26f=915;ls26l=909;ls27f=911;ls27l=900;ls28f=918;ls28l=910;ls29f=909;ls29l=896;ls2f=903;ls2l=915;ls30f=899;ls30l=902;ls31f=921;ls31l=910;ls32f=943;ls32l=984;ls33f=940;ls33l=901;ls34f=904;ls34l=931;ls35f=899;ls35l=905;ls36f=907;ls36l=910;ls37f=928;ls37l=907;ls38f=905;ls38l=908;ls39f=911;ls39l=906;ls3f=909;ls3l=925;ls40f=902;ls40l=923;ls41f=931;ls41l=904;ls42f=909;ls42l=911;ls43f=910;ls43l=909;ls44f=908;ls44l=994;ls45f=899;ls45l=998;ls46f=894;ls46l=922;ls47f=911;ls47l=911;ls48f=927;ls48l=931;ls49f=905;ls49l=931;ls4f=910;ls4l=966;ls50f=889;ls50l=899;ls51f=907;ls51l=923;ls52f=908;ls52l=905;ls53f=912;ls53l=891;ls54f=903;ls54l=905;ls55f=896;ls55l=920;ls56f=912;ls56l=931;ls57f=907;ls57l=887;ls58f=892;ls58l=947;ls59f=900;ls59l=954;ls5f=925;ls5l=897;ls60f=908;ls60l=917;ls61f=916;ls61l=947;ls6f=906;ls6l=905;ls7f=913;ls7l=939;ls8f=913;ls8l=997;ls9f=908;ls9l=898;rf10p=349;rf10v=720;rf11p=333;rf11v=686;rf12p=374;rf12v=715;rf13p=381;rf13v=997;rf14p=358;rf14v=665;rf15p=400;rf15v=738;rf16p=431;rf16v=698;rf17p=422;rf17v=645;rf18p=439;rf18v=654;rf19p=422;rf19v=680;rf1p=320;rf1v=676;rf20p=412;rf20v=682;rf21p=379;rf21v=697;rf22p=354;rf22v=995;rf23p=377;rf23v=668;rf24p=363;rf24v=750;rf25p=400;rf25v=758;rf26p=378;rf26v=662;rf27p=376;rf27v=691;rf28p=361;rf28v=777;rf29p=391;rf29v=756;rf2p=329;rf2v=999;rf30p=388;rf30v=684;rf31p=397;rf31v=739;rf32p=383;rf32v=707;rf33p=359;rf33v=999;rf34p=355;rf34v=735;rf35p=312;rf35v=722;rf36p=336;rf36v=659;rf37p=345;rf37v=718;rf38p=480;rf38v=732;rf39p=395;rf39v=674;rf3p=387;rf3v=294;rf40p=411;rf40v=000;rf41p=400;rf41v=718;rf42p=387;rf42v=682;rf43p=397;rf43v=786;rf44p=415;rf44v=999;rf45p=404;rf45v=999;rf46p=402;rf46v=694;rf47p=395;rf47v=797;rf48p=379;rf48v=755;rf49p=422;rf49v=660;rf4p=389;rf4v=714;rf50p=406;rf50v=675;rf51p=408;rf51v=595;rf52p=431;rf52v=683;rf53p=419;rf53v=733;rf54p=481;rf54v=753;rf55p=443;rf55v=724;rf56p=420;rf56v=761;rf57p=353;rf57v=773;rf58p=321;rf58v=661;rf59p=000;rf59v=669;rf5p=355;rf5v=655;rf60p=361;rf60v=690;rf61p=299;rf61v=634;rf6p=386;rf6v=999;rf7p=344;rf7v=742;rf8p=308;rf8v=996;rf9p=307;rf9v=947;s10f=900;s10l=900;s10r=909;s11f=945;s11l=917;s11r=990;s12f=897;s12l=906;s12r=897;s13f=901;s13l=893;s13r=999;s14f=898;s14l=901;s14r=906;s15f=900;s15l=909;s15r=999;s16f=911;s16l=914;s16r=903;s17f=908;s17l=898;s17r=895;s18f=900;s18l=908;s18r=926;s19f=929;s19l=910;s19r=893;s1f=997;s1l=768;s20f=918;s20l=909;s20r=917;s21f=898;s21l=898;s21r=880;s22f=935;s22l=914;s22r=925;s23f=901;s23l=906;s23r=937;s24f=773;s24l=912;s24r=521;s25f=918;s25l=885;s25r=912;s2f=971;s2l=991;s2r=913;s3f=900;s3l=688;s3r=895;s4f=907;s4l=856;s4r=929;s5f=914;s5l=985;s5r=895;s6f=912;s6l=910;s6r=900;s7f=888;s7l=907;s7r=926;s8f=890;s8l=917;s8r=891;s9f=907;s9l=911;s9r=999;tantalumrodr=000;tantalumrodz=432;
1.474123 (2495.6 Mpts) [v4.41b] <ChicaneLinacB> #time=48566; by AudioElf
)

Perhaps, your right, Stephen should provide a "standard" benchmarking program, and normalize scores between program versions and optimizations :_P (smilie intentional)

RGtx
2010-01-09 21:57:41
not :_P
[TA]Assimilator1
2010-01-10 10:15:23
Yea you gotta leave the underscore out (as I mentioned in my smilie thread).

If theirs that bigger a difference then benchmarking is almost pointless atm! 
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io RSS feed

Site has had 25162321 accesses.