Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 3. (Read 2347588 times)

I use the newest drivers. I have a Radeon Pro Duo in 4 different systems because none of the miners work right if I put 2 of them in 1 system. So I can install either the Pro drivers or the Adrenaline drivers. It doesn't make any difference however with Sp4.

i have the same problem with GCN 1 & 2 i told SP about it,a few minutes into mining it starts submitting shares with a high ping due to lag, in my case even SP3 lags, what driver version are you using ?

Couple issues with SP4.

Doesn't work with R9 280(x)'s anymore. Crashes the driver as soon as they start up. They work fine with SP3.

Insane lag a few minutes into mining on any of my systems.

I have been using SP3 for about a month now because of these issues.

Fee problems(the fee is mined in the first hour of mining):

mine 1 hour and quit: fee 12%
mine 2 hours and quit: fee 6%
mine 3 hours and quit: fee 3%
mine 4 hours and quit: fee 1.5%
mine 12 hours and quit: fee 1%
...

So if your rig is unstable and crash, restart every hour, 12% fee...


Gpu search high intensity problems

1. Searching 500 000 solutions in a low difficulty. 17 solutions found. spminer will only report 2 of them.
2. Searching 500 000 solutions and the pool change difficulty and reject your work.(stale share) you loose 500khash
3. Searching 500 000 solutions and the pool change block and reject your work. you loose 500khash
4. Searching 500 000 solutions and the pool restart your work. you loose 500khash (nicehash, zergpool and other multipools does this all the time)
5. Searching 500 000 solutions and the pool have a new job. you loose 500khash
6. Searching 500 000 solutions instead of 10 000 require more gpu memory and more power. Make sure that your cards don't trottle and and feed them enough power.

Spminer is faster in finding blocks.
 www.miningpoolhub.com and www.hashalot.net rewards the blockfinders and doesn't trow away the work in 2-5 that get rejected on other pools. As a miner you want to get payed for all your work.

Mining on a fixed low difficulty will hurt pool performance because solutions are skipped.

The faster eth miner in this case is not that much important ( unless its really faster and its a different story).
Its more important to minimize the switch time between VTC and ZIL.
So, for the 4GB card, it means to create the Zil dag in memory without loosing to much in speed during VTC time but being ready when zil time comes.

yes it could be, when you drop the next release with the fix i'll tinker around with the -i command & see,
thank you

The good old 7970 and rx 280. The 7970 is from 2013. It might be that the default intensies are too high for this card, so in the next version I have fixed the -i parameter. I don't use these cards anymore so cannot test it. But I will run some pool test on my rx 570/580 cards.

-AMD,7 old Gen 1 & 2 GCN cards
-i hopped on all vertcoin pools same result
-i'm not dual mining
-i tried all spminer releases same result
-my rig (an 1150 platform) can mine Primecoin & Turtlecoin just fine,two of the hottest most stressful algos out there
-as i said VerthashMiner works fine too
-the Driver is 18.6.1
-Ram 8 GB
 

just using the basic ....SPMiner -o stratum+tcp://xxxxx -u xxxxxx -p x --verthash-data verthash.dat --all-cl-devices

Amd or Nvidia? Whats the specs on the rig. Wich pool. Are you dual mining with zil? Wich version of the spminer?. Whats your launch config? Whats the corclock, memory clock and tdp of your cards? Driver version?

i have the same issue too i suspect the default intensity of the miner is either too high or the miner itself is unstable making the entire system unstable & causing high ping'd submitted shares,

for example i can run the web browser for a quick search while mining with any miner except SPminer which behaves as if it is very taxing on the hardware even by itself

because of this & with less hashrate i make more with VerthashMiner

The ping is not high, it's the program that display incorrect results.. Will try to fix it in the next release

Okay I see.

We ran into some issue with SPminer on the p2pool after a while as well. For some reason after a long time the ping goes crazy high.

I did a lot of trouble shooting, and the only thing that fixed it was using a different miner.

The issue is strange though, it only seems to happen with spminer if I am running a p2pool on the same network? I tried connecting to hashalot as well and ended up with the same high ping after a while.

If I use SPminer without p2pool running on the same network there is no issue, I am so confused?

Setup is p2pool and one miner on one machine, then another miner on a different machine, pointing to the local p2pool.

The same setup running with VHM instead works with no issues.

Maybe I will try moving the pool to a different network, or try with miningcore instead.

I think the selection of pool is important here. For spminer Miningpoolhub.com and Hashalot.net give the best returns. Up to 10% bether than the other pools.

---

This option is only for pools that give a penalty for stale/invalid shares. F.ex the Supernova pool. Every time you see a skipped solution, you loose hashrate poolside. Supporting p2p pools is difficult because of the rapid difficulty changes and low difficulty. It requires tuning and codechanges to perform well. For low difficulty you usually need low intensity. A high intensity meens that you give the gpu alot of solutions to search for in each bulk and this can take a few seconds to complete. This approach is good for solomining and high difficulty, but not good for rapid difficulty changes and pools with low diff.

---

Poolside +10% faster than the Team Red and the SRB miner...

First tests with the ethereum miner show promising results. A few percent faster than the competition.. More profit in the pockets of the miners.

---

The fee normally mined in the first few minutes of mining, then you get around 12 hours of mining without fee, and the fee repeats. Good results with Long term mining and rigs that doesn't crash. The competion is mining a few shares every hour here and there.

[moderator's note: consecutive posts merged]

Nice. ETC and ZIL is good.

Can you please add this .bat file to future releases? https://asciidity.xyz/start_acidpool.bat

KEEP IN MIND THAT ETH MINING STILL PAYS BEST--

The new GPUs with more than 4GB VRAM will mine ETH.  Miners with older, smaller cards need software that runs on older systems.  The algorithms that you want to incorporate in your miner are just fine for the smaller cards.

Ethereum Classic (ETC) had a spike last week that had it paying second only to ETH.  Not too long ago, before it forked, it paid better.  I just need software that is going to run on my Linux rigs.  VTC is my top earner for my smaller cards, but nVidia is left out.  Software designed for the latest and greatest isn't running on smaller GPUs that fit the algos (VTC, ETC, ZIL) that need less VRAM.

--scryptr

The next update wil include zilminer, ethereum, ethereum classic. This will take some time..

t-rex won't work with that driver version either. Needs 46x.xx+.

---

This gives and unknown option error in spminer 4 with cuda 10? I also tried --skip-submit-stales like in the readme?

EDIT: it's --skip-stale-work

[moderator's note: consecutive posts merged]

TEST RESULTS--

I ran spminer with the flag "-h".  This is what I got:

 "./spminer_cuda10.1: error while loading shared libraries: libcudart.so.10.1: cannot open shared object file: No such file or directory" .

Thanks for looking into things.  I have tried to update the driver in the past and failed.  I may try again, but I don't want to hose up a working system with a failed install.

--scryptr

Can you type the excact error message please?

NVIDIA-SMI RESULTS--

The driver version is 430.26, with CUDA 10.2.       --scryptr