Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 1173. (Read 2347597 times)

Yes I get the FFFFFFF one when I was doing my tests.  I assumed because of all the FFFFFF each time it was not a valid solution.

(Sorry if my FFFFFFF are not the right number of F.)

Thanks for testing.
Seems to be an error in the test for this nounce value 0xffffffff. I think i need to change the test. This will not reduce the speed of the miner.

The miner should find blocks now, but if you get invalid nounces except 0xffffffff then the hashspeed is broken.

Tnx for testing. 750ti cards do good with multiple of 5. 980 and 960 with multiple of 8 And 970 with multiple of 13.

Default in release 4 is :

750ti: 5
970: 13
980/960: 16

Release 4 is around 2640 khash on a reference 960. Standard clocked.

Try to run it  for a while in benchmark mode( -benchmark). Do you get any "does not validate on the cpu?"

-X factor on 750 ti test from 1 to 25 (26 is out of memory)

My Rig

6 EVGA 750 Ti SC cards -No overclock -no memory overclock -no bios mod -no extra power plug
Windows 8.1 Pro
Nvidia 347.25 Driver
Spreadminer4

I did a comparison of -x 1 to 25 for about 2 minutes and looked at the high and low and eyeball average (No measurement taken)
Got me down to testing 6 settings
Base (No -x) 1738
-x 5  1739
-x 10 1732
-x 15 1737
-x 20 1742
-x 21 1739

Did a timed test of 2 minutes (Throwing out the start up hash rates, first 6 readings at sub 1600 Kh/s speeds). Used Excel to get average numbers below.
Base (No -x) 1730
-x 5  1731
-x 10 1737
-x 15 1739
-x 20 1738
-x 21 1730

So I removed -x21 because why push the card that hard if it does not give good results.

I did a 5 minute test then, again throwing out the startup hashes below 1600 Kh/s. (See if heat slowdown would be an issue.) Used Excel to get average numbers below.
Base (No -x) 1726.54
-x 5  1728.48
-x 10 1732.68
-x 15 1737.46
-x 20 1736.99

So the winner for my Rig is -x 15

Here is all of the results from first test. It surprised me how well -x 5 did

        Min   Max   Eyeball Avg
Base   1697   1750   1738
-x 1   1380   1470   1450
-x 2   1504   1576   1553
-x 3   1549   1595   1572
-x 4   1549   1593   1578
-x 5   1699   1755   1739
-x 6   1671   1718   1692
-x 7   1545   1602   1586
-x 8   1702   1750   1724
-x 9   1629   1691   1668
-x 10   1689   1755   1732
-x 11   1680   1737   1710
-x 12   1678   1735   1707
-x 13   1673   1722   1700
-x 14   1577   1709   1684
-x 15   1705   1755   1737
-x 16   1690   1745   1720
-x 17   1687   1733   1720
-x 18   1681   1731   1702
-x 19   1676   1722   1716
-x 20   1707   1761   1742
-x 21   1699   1750   1739
-x 22   1694   1749   1726
-x 23   1692   1743   1721
-x 24   1241   1727   1589
-x 25   650   1758   1400
-x 26 Out of Memory Crash

sp_, release 4 still broken, hash problems.

Thanks for that!

This is all the dll I have in that folder
cudart32_65.dll
msvcr120.dll
pthreadVC2.dll

Maybe it is another one of them??? Just guessing

Considering all cards slightly overclocked with tsiv´s 0.1r2 miner:

750Ti: 1,7MH/s
960GTX: never tested, I believe it´s about 2x750Ti = 3,0-3,4MH/s
970GTX: 4,0MH/s
980GTX: 5,3 MH/s

Cheers

what are the typical hashing speeds for the Spreadcoin Algo for 750, 960, 970 and 980?

hi, can anyone help me with crashing driver only with 970 gtx gpus while mining with ccminer all realases,  tried all 3 different drivers on win 8.1 gpus are not overclocked and temps are great, they are working fine while stress testing or gaming. 5xx and 6xx and 7xx series working fine with tsiv/tpruvot older miners and 770gtx works fine with sp releases only 970 gtx are crashing randomly sometimes 4 times in one hour and sometimes once in a 4 hour etc, is that normal, i tried reinstalling everything milion times but have no luck pls help  

i am mining mjollnir and i tried all releases from 1.5.1 upwards to sp 32  

Sorry, welcome and good luck! Feel free to pm me if you need help.

<--- look at my stats, I'm a newbie. This is the 1st day. I will try single algos and see what difference I get, my wallet is not synced yet. I'd assume that the pool calculates what profitable and switches you to it. At the moment its X14 (alipaycoin) and quark (a mixture) no idea they really are the best, though I see many more people are on X11, going to try a few different things over the next few days and see what happens.

Does that actually grant you higher payouts at the end of the day? I'd imagine constant switching is just lagging behind what was profitable. And the more people use a script like it the worse it gets.

Edit: CPU validation errors are still there with v4.

I'm running a script that cycles through different algos on YAAMP so that in theory the pool gets me to mine the most profitable using  "-p x11,x13,x14,x15,quark,lyra2"  to tell the pool what I can do
earlier today I was getting lots of rejects on quark, sometimes down to 10% accept rate main reason was "job not found", now its much better, 90% accept and higher,  the rejects are "duplicate share" so I assume the earlier problem was not down to my miner but more a fault of the pool/network that has been fixed.
 The only other algos YAAMP wanted me to mine were X13 and X14 the others were skipped, I get less rejects with those.
Now it still disconnects with stratum_recv_line failed every so often on quark ,5-30 mins range, skips all the other algos and goes right back to quark. I might as well just run quark only to get rid of the delays induced by failing the others.

I also had 344.75 and was getting pretty much 100% crap. Upgraded to 347.25 and running with -x 13:



Not exactly great but seems a bit better. Trying to get hold of CUDA 7.0 to try it out, might be a while.

Also made the found block notification a lot less noisy in the latest release, that huge dump was intended for debugging and I just didn't realize it was getting printed even when not running with -D.

#3:



edited

Same thing here!  Maybe you are right.  From 10-12 miners a few weeks ago...        --scryptr

I'd say it's the driver, but I don't think I've mined spread with the older drivers and I had no rejects until a couple days ago. I set my 2 970s back to default instead of -x 13, but so far no rejects or accepts.  Had 1 earlier when I had it at 13.  Just shooting in the dark now though.