GPU to 72 status... - Page 37

chalsall · 2012-01-27, 21:14

Just a quick note. As of a little while ago the project has found 1,391 factors.

LL/DC tests saved: 2,500.
GHz Days saved: 213,629 (602 GHz YEARS)!!!

Dubslow · 2012-01-27, 23:05

Ya know, even if nothing on the site changes, it would still be interesting to compile a list of all the ratios for everybody. Just like in a CSV or something.

chalsall · 2012-01-27, 23:27

Quote:

Originally Posted by Dubslow

Ya know, even if nothing on the site changes, it would still be interesting to compile a list of all the ratios for everybody. Just like in a CSV or something.

Could you give me an example of what kind of data you would like exposed, and I will do my best to facilitate?

And here's an exercise for you from the other thread: please express your request in SQL and/or C (or, of course, Perl).

Dubslow · 2012-01-27, 23:50

http://gpu72.com/reports/worker/829a...ca0453117d491/

My ratio is 30 there.

For our good friend Craig Meyers, the ratio is 500.

So for us 2,

Code:

30,500

would be the start of it.

Of course, the main problem (that only just occurred to me now) is that the graphs have multiple work types for each hardware type, so unless you calculate what the scale would be by only using a GPU work line and a CPU work line, this data is in fact less interesting than I thought it might be.

Assuming that the top axis-G label number for worker i is scale[i][0] and that the axis-C number is scale[i][1], then the code might be:

Code:

#std includes
int main () {

int i;      
for(i=0; i<N; i++){                         // N is number of workers
   printf("%d,", scale[i][0] / scale[i][1] );
}
}

Barring the extra comma, that's what I'm looking for.

Edit: Typing an italic i is incredibly weird looking.

chalsall · 2012-01-28, 00:11

Quote:

Originally Posted by Dubslow

Of course, the main problem (that only just occurred to me now) is that the graphs have multiple work types for each hardware type, so unless you calculate what the scale would be by only using a GPU work line and a CPU work line, this data is in fact less interesting than I thought it might be.

LOL. That is one of the interesting things about SW development; sometimes what you think might be interesting isn't, and what you think isn't is...

Quote:

Originally Posted by Dubslow

Assuming that the top axis-G label number for worker i is scale[i][0] and that the axis-C number is scale[i][1], then the code might be:

But you've taken a short-cut (used English), and expected me to fill in the code.

How do I calculate the values for the two-dimensional (or, perhaps, three-dimensional?) array?

(As an aside, I know the answer. But I am interested to see if you can tell me. I believe nothing is better for learning than what I call "driving problems".)

chalsall · 2012-01-28, 01:02

Is this something like what you are after (one line of SQL)?

Code:

+----------------------------------+------------------+------------------+------------------+-----------------+-----------------+-----------------+
| User                             | Total            | DCTF             | LLTF             | P1              | DC              | LL              |
+----------------------------------+------------------+------------------+------------------+-----------------+-----------------+-----------------+
| fc9f090d094ad7c7eff10a39caffe3a4 | 93052.7561291456 |     0.0000000000 | 92219.4257501364 |  453.2804408073 |  380.0499382019 |    0.0000000000 | 
| 7e6a2e592a37a719fac4f765eb0f6ca8 | 78384.8326845169 |   891.8903484344 | 76810.1566689014 |  682.7856671810 |    0.0000000000 |    0.0000000000 | 
| b30b0ed22bbd92716af40ebf12fbf217 | 29572.6529233456 | 23220.2679257393 |  5253.4110212326 |  461.7658102512 |  129.7618846893 |  507.4462814331 | 
| 773ea4ea4e2b84a4c8473454655ac6da | 29168.0139116049 |   305.5261943340 | 28350.9771536589 |  351.9818370342 |   87.0477237701 |   72.4810028076 | 
| 19437658c4ba52abaa8aacbb8da29007 | 28462.3662837744 |     2.2766087055 | 28059.0157786608 |    0.0000000000 |  111.5592327118 |  289.5146636963 | 
| b49e80e9382626afc7ebe2f81662dadb | 19590.9275262356 |  3762.2742357254 | 14269.2052748203 | 1147.2698254585 |  267.4067897797 |  144.7714004517 | 
| 390a11e492481f62d201b6578395bd14 | 18707.7199903727 |   121.6711617708 | 17404.5658966303 |  674.6213215590 |    0.0000000000 |  506.8616104126 | 
| 39ca3b02920ccc839ea59d34b6ea6128 | 17809.8644920588 |     3.5865688324 | 16208.4802600145 | 1076.9546608925 |  520.8430023193 |    0.0000000000 | 
| d822948f629fc0203f56ff452665fed9 | 16473.2675839663 |  3525.5860016346 | 12608.8956726789 |  124.6123542786 |  141.5349826813 |   72.6385726929 | 
| e60fb101e5d8bab038a0ef48531126fb | 16471.9752160311 |  1437.0388631821 | 14929.6735153198 |   83.7873393297 |   21.4754981995 |    0.0000000000 | 
| 89edb735f68ff3faa688634ac776d5f8 | 15810.2953671217 | 15665.9857981205 |   144.3095690012 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| 2423ae6e8f696d5e7d1447de91ca35a6 | 14553.1118537188 |  2359.6173614264 | 10584.5242302418 |  103.9309172630 | 1215.4569854736 |  289.5823593140 | 
| 8ad778c87c3d88f688fef4ebad36994c | 12818.6088447571 |   647.4306507111 | 11516.4009673595 |  539.0026142597 |   43.3852691650 |   72.3893432617 | 
| 5b6f7bd5a6b6e7e6988c81e70d6cd07d | 11774.6411081553 |   281.2115180492 |  9311.3107892275 |  794.2781713009 |  152.9845428467 | 1234.8560867310 | 
| 5d526244a9c705e5d5958ec8b762e44a | 11400.9555712938 |     0.0000000000 | 10761.8047977686 |  639.1507735252 |    0.0000000000 |    0.0000000000 | 
| 15aa325b9d9ee716c825070059f2337a | 10728.3232730627 |  3193.5058393478 |  7106.5047348738 |  211.4462742805 |    0.0000000000 |  216.8664245605 | 
| 7b09336e7e884823d38ee0422b96489e |  9313.8079361916 |  2503.7048001289 |  3460.4042844772 | 2040.0348482132 | 1309.6640033722 |    0.0000000000 | 
| 74fa061571f40ebf723c1b19765143a2 |  7675.5299279690 |     0.0000000000 |  6365.9617524147 |  364.1169044971 |    0.0000000000 |  945.4512710571 | 
| 56f1b7572536a14513b08c88b2ba9578 |  7368.3487591743 |     2.2235989571 |  6273.8510737419 |  852.5724205971 |  239.7016658783 |    0.0000000000 | 
| c298d077913beb302bd98e0f1ff3623a |  6149.5275577307 |    35.2031968832 |  5657.1156709194 |    7.3414528370 |   86.8459205627 |  363.0213165283 | 
| 829a683f5d991e17d4cca0453117d491 |  6058.8373221159 |    60.5250223875 |  5380.1917250156 |  205.0804498196 |  195.4776630402 |  217.5624618530 | 
| 82ee93ef1cfd803387fbcd441eb0c0db |  5735.7941768169 |  2228.3989956379 |  3388.7941403389 |   96.9764451981 |   21.6245956421 |    0.0000000000 | 
| aaa579791891591b0fb791aca5a8f9f7 |  5598.1223777533 |     0.0000000000 |  5598.1223777533 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| 89ef632e1874ea7cff7db1f47650c445 |  4721.8471994400 |     0.0000000000 |  4721.8471994400 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| a61d114e14e94c21a85ec3cf6b2ea883 |  4665.6311793327 |     0.0000000000 |  4665.6311793327 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| d2720bb3b549c17449b37fe44e0132a4 |  4548.5567498207 |     0.0000000000 |  4285.7440457344 |    0.0000000000 |   44.6570720673 |  218.1556320190 | 
| 8f15297ed607932fd2e1d0d23d4880c0 |  4380.6567510366 |   509.4930229187 |  3809.6007031202 |   61.5630249977 |    0.0000000000 |    0.0000000000 | 
| 4d89b8ff781e27a0fe80450cd4cd74b6 |  3996.0682355165 |   322.2086298466 |  3670.5757766962 |    3.2838289738 |    0.0000000000 |    0.0000000000 | 
| b87b5fa740c4e5002bd21b6d06e93a6a |  3457.1506243944 |    41.5396325588 |  3322.4374296665 |   93.1735621691 |    0.0000000000 |    0.0000000000 | 
| d9c51d91f33f940d4aa8b75597c7baa4 |  3138.4686372280 |    35.1989448071 |  2916.5600125790 |  186.7096798420 |    0.0000000000 |    0.0000000000 | 
| 659a202c7e9ff182479bb0a2d32eed40 |  2652.9638943672 |    22.8949916363 |  2596.4105024338 |   33.6584002972 |    0.0000000000 |    0.0000000000 | 
| 357ac6da059edbbe017dc4cf8ba6cb8f |  2459.6024336815 |  1332.3776972294 |   793.0798296928 |  106.9247052670 |  154.9639453888 |   72.2562561035 | 
| 412dcee751476f63b72b20e66a2cd03e |  2427.4980827570 |     0.0000000000 |     0.0000000000 | 1622.4985271692 |  804.9995555878 |    0.0000000000 | 
| 7fe2e567d20835388ee8c0428e4efd04 |  1548.4534907341 |   114.2741448879 |  1404.6411375999 |   29.5382082462 |    0.0000000000 |    0.0000000000 | 
| 4fed06b8b93c3bd30502dd0674b94a49 |  1452.9644484520 |     0.0000000000 |  1452.9644484520 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| a1103bbec5450d20ce65569281886283 |  1037.4996016026 |   840.9731371403 |     0.0000000000 |    0.0000000000 |  196.5264644623 |    0.0000000000 | 
| 680b8b282d0fd112d34878998c201d9c |  1011.8093857765 |     2.3912599087 |     0.0000000000 |  157.4539859295 |  198.2143306732 |  653.7498092651 | 
| bd25eec0e992020077c0a9944ad678ad |   946.3190679550 |   129.3302781582 |   814.5159578323 |    2.4728319645 |    0.0000000000 |    0.0000000000 | 
| 799ebc472212227f0fd88cff837004bf |   802.0690443516 |    86.4216604233 |   715.6473839283 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| 03f697129bd9ea56e3300957ab12093a |   626.7700088024 |     2.0425975323 |   624.7274112701 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| 1ef966aca8d632d34f364f7713f24df2 |   614.5606753826 |     0.0000000000 |     4.9073691368 |   76.5187942982 |  387.8962535858 |  145.2382583618 | 
| de60723d06bc1a86a359daa4755196d6 |   605.1823937893 |     0.0000000000 |    20.5088188648 |  584.6735749245 |    0.0000000000 |    0.0000000000 | 
| 8cf99e218e800a2c4ee3c7e3699d9393 |   584.6152449846 |     0.0000000000 |     0.0000000000 |  207.6526175737 |   86.4503669739 |  290.5122604370 | 
| 08edf8256558240deccdd67185a72419 |   570.1444544792 |     0.0000000000 |     0.0000000000 |  352.5807261467 |    0.0000000000 |  217.5637283325 | 
| d8a75f85f90457298bd3c366a8de2410 |   569.4388532639 |     0.0000000000 |     0.0000000000 |   63.5871133804 |  505.8517398834 |    0.0000000000 | 
| 143471536f544b17590aad4023de7132 |   518.8536686897 |     0.0000000000 |     0.0000000000 |  453.6173176765 |   65.2363510132 |    0.0000000000 | 
| 2877aa0e1c37b72ea13b99ae1f87e5f9 |   376.9706678391 |   149.1084997654 |   216.6685962677 |   11.1935718060 |    0.0000000000 |    0.0000000000 | 
| d0a0a9610115a8227c686f0d8951998b |   359.4586672783 |   140.1771779060 |   219.2814893723 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| fa50efecb5afbe03f29823943d4d2030 |   304.1713733673 |     0.0000000000 |   304.1713733673 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| f41a55635b59eed28f9641feb3d0c625 |   130.9898891449 |     0.0000000000 |     0.0000000000 |    0.0000000000 |  130.9898891449 |    0.0000000000 | 
| 42b0b8e7827e0255d58d5e8187f57ce2 |    21.4284191132 |     0.0000000000 |     0.0000000000 |    0.0000000000 |   21.4284191132 |    0.0000000000 | 
| e1a1a13121a9cb9447e66ba643dd2323 |    10.5839843750 |     0.0000000000 |    10.5839843750 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| c6c26365859ffecedf398b71c8828ac8 |     9.3763933182 |     2.1354253292 |     5.0018186569 |    2.2391493320 |    0.0000000000 |    0.0000000000 | 
| c99dcc0fe7afca8be9e45471dc2acbe6 |     8.7402048111 |     0.0000000000 |     8.7402048111 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| 09dc14d290a4700bb74e500395e92d8e |     5.2093033791 |     0.0000000000 |     0.0000000000 |    5.2093033791 |    0.0000000000 |    0.0000000000 | 
| 0d9dc47515b25a7ab0d77ed071be43bf |     4.6094813347 |     0.0000000000 |     4.6094813347 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| 3a43a8d272835f23bce1b55bb0e1b842 |     3.6444642544 |     0.0000000000 |     0.0000000000 |    3.6444642544 |    0.0000000000 |    0.0000000000 | 
| 5067bf2168d1e4f989287acf7c8e13f5 |     2.1243565083 |     2.1243565083 |     0.0000000000 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
| d43207809b2f4cf690f29ba2e8a9288a |     1.4474825859 |     0.0000000000 |     1.4474825859 |    0.0000000000 |    0.0000000000 |    0.0000000000 | 
+----------------------------------+------------------+------------------+------------------+-----------------+-----------------+-----------------+

Bdot · 2012-01-28, 02:05

Hmm, the LLTF and LL columns have the same content ... was this to test if anybody notices?
Edit: And now you corrected it before my post was ready!

I guess Dubslow was after something like

Code:

Select USER, (LLTF+DCTF)/(P1+DC+LL) from wherever_you_store_that_stuff;

Kind of a ratio between GPU and CPU work done. However, if he claims the result for him was 30, then it is probably something like

Code:

Select USER, MAX_OF(LLTF,DCTF)/MAX_OF(P1,DC,LL) ... ;

Where MAX_OF is shall be some function that returns the maximum value of its arguments (I do not mean MAX() which returns the maximum value of a whole column in SQL). Not sure if such a function already exists in SQL, otherwise perl needs to take care of the "calculation".

Would these listings be interesting? Certainly not.

And yes, you may need to take care of the special case that someone has no CPU work whatsoever ...

flashjh · 2012-01-28, 03:48

Quote:

Originally Posted by chalsall

Hey all.

I've made a small change to the individual worker's graphs

Awesome changes, thanks.

chalsall · 2012-01-28, 04:49

Quote:

Originally Posted by Bdot

Where MAX_OF is shall be some function that returns the maximum value of its arguments (I do not mean MAX() which returns the maximum value of a whole column in SQL). Not sure if such a function already exists in SQL, otherwise perl needs to take care of the "calculation".

It's a little convoluted, but it can be done in SQL. For one side of the equation:

Code:

select User, if(sum(DCTF) > sum(LLTF), sum(DCTF), sum(LLTF)) ... from Assigned group by User;

And SQL will return NULL in the case of division by zero.

Dubslow · 2012-01-28, 19:20

My links were to the graphs. The reason I shortcutted and used English is because I have no idea how you calculate the scale on worker's graphs. Look at my chart:
http://gpu72.com/reports/worker/829a...ca0453117d491/

On the left, my top GPU axis label is 7500 GD, whereas on the right, my top CPU axis label is 250 GD. Diving these two produces a ratio of 30 from the CPU axis scale to the GPU axis scale. (Of course, like I said, what's more correct is if you produce the same graph except with total CPU and GPU work as lines, as opposed to each worktype getting its own line.) I have absolutely no idea how to use SQL, I only barely know what it is, and how you produce the graphs and their axis scalings is way beyond me.

chalsall · 2012-01-28, 19:54

Quote:

Originally Posted by Dubslow

The reason I shortcutted and used English is because I have no idea how you calculate the scale on worker's graphs.

But you do know the variables involved. There are only five of them (six including the total) indexed by User.

Quote:

Originally Posted by Dubslow

On the left, my top GPU axis label is 7500 GD, whereas on the right, my top CPU axis label is 250 GD. Diving these two produces a ratio of 30 from the CPU axis scale to the GPU axis scale. (Of course, like I said, what's more correct is if you produce the same graph except with total CPU and GPU work as lines, as opposed to each worktype getting its own line.)

But that would only hide (by way of vertical axis compression) the resolution of the individual worktypes' lines.

(And my understanding from your comments was you wanted all the individual graphs to use the same scales. Please let me know if I misunderstood you.)

Please understand that I take everyone's feedback seriously.

But another part of software development is knowing that not every user's comment or feedback makes sense. (Most do; and all should be considered (particularly "This is strange..." type comments). But not all.)

IMHO, unless you can take the raw data I provided to you above from the SQL database (import it into a spreadsheet, or filter it through a RegEx and import it into a two dimensional array) and tell me how this can be used to improve the graphs for everyone, this particular suggestion doesn't make sense to me.

2012-01-27, 21:14	#397
chalsall If I May "Chris Halsall" Sep 2002 Barbados 2627₁₆ Posts	Milestone: 2,500 LL/DC tests saved!!! Just a quick note. As of a little while ago the project has found 1,391 factors. LL/DC tests saved: 2,500. GHz Days saved: 213,629 (602 GHz *YEARS*)!!!

2012-01-27, 23:50	#400
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 1C35₁₆ Posts	http://gpu72.com/reports/worker/829a...ca0453117d491/ My ratio is 30 there. For our good friend Craig Meyers, the ratio is 500. So for us 2, Code: 30,500 would be the start of it. Of course, the main problem (that only just occurred to me now) is that the graphs have multiple work types for each hardware type, so unless you calculate what the scale would be by only using a GPU work line and a CPU work line, this data is in fact less interesting than I thought it might be. Assuming that the top axis-G label number for worker i is scale[i][0] and that the axis-C number is scale[i][1], then the code might be: Code: #std includes int main () { int i; for(i=0; i<N; i++){ // N is number of workers printf("%d,", scale[i][0] / scale[i][1] ); } } Barring the extra comma, that's what I'm looking for. Edit: Typing an italic i is incredibly weird looking. Last fiddled with by Dubslow on 2012-01-27 at 23:51

2012-01-28, 02:05	#403
Bdot Nov 2010 Germany 3×199 Posts	Hmm, the LLTF and LL columns have the same content ... was this to test if anybody notices? Edit: And now you corrected it before my post was ready! I guess Dubslow was after something like Code: Select USER, (LLTF+DCTF)/(P1+DC+LL) from wherever_you_store_that_stuff; Kind of a ratio between GPU and CPU work done. However, if he claims the result for him was 30, then it is probably something like Code: Select USER, MAX_OF(LLTF,DCTF)/MAX_OF(P1,DC,LL) ... ; Where MAX_OF is shall be some function that returns the maximum value of its arguments (I do not mean MAX() which returns the maximum value of a whole column in SQL). Not sure if such a function already exists in SQL, otherwise perl needs to take care of the "calculation". Would these listings be interesting? Certainly not. And yes, you may need to take care of the special case that someone has no CPU work whatsoever ... Last fiddled with by Bdot on 2012-01-28 at 02:07

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Status	Primeinator	Operation Billion Digits	5	2011-12-06 02:35
62 bit status	1997rj7	Lone Mersenne Hunters	27	2008-09-29 13:52
OBD Status	Uncwilly	Operation Billion Digits	22	2005-10-25 14:05
1-2M LLR status	paulunderwood	3*2^n-1 Search	2	2005-03-13 17:03
Status of 26.0M - 26.5M	1997rj7	Lone Mersenne Hunters	25	2004-06-18 16:46

2012-01-27, 23:05	#398
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 1C35₁₆ Posts	Ya know, even if nothing on the site changes, it would still be interesting to compile a list of all the ratios for everybody. Just like in a CSV or something.

2012-01-28, 19:20	#406
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 16065₈ Posts	My links were to the graphs. The reason I shortcutted and used English is because I have no idea how you calculate the scale on worker's graphs. Look at my chart: http://gpu72.com/reports/worker/829a...ca0453117d491/ On the left, my top GPU axis label is 7500 GD, whereas on the right, my top CPU axis label is 250 GD. Diving these two produces a ratio of 30 from the CPU axis scale to the GPU axis scale. (Of course, like I said, what's more correct is if you produce the same graph except with total CPU and GPU work as lines, as opposed to each worktype getting its own line.) I have absolutely no idea how to use SQL, I only barely know what it is, and how you produce the graphs and their axis scalings is way beyond me.