Ninjastic
PostEdited versionsQuotes to this post
Post
Topic
Board Mining (Altcoins)
Re: NPlusMiner - GUI Pool/Algo switching miner with Plus logic (NVidia)
by
MrPlus
on 17/03/2018, 18:06:44 UTC
Quote from: JCS-CCT on March 17, 2018, 05:39:54 PM
Quote from: Just_a_miner on March 17, 2018, 05:24:04 PM
Quote from: JCS-CCT on March 17, 2018, 04:48:13 PM
Quote from: Just_a_miner on March 17, 2018, 11:34:46 AM
Quote from: JCS-CCT on March 16, 2018, 04:23:48 AM
Quote from: skablast on March 16, 2018, 12:30:13 AM
with hsr ? 1050ti 4 gb

We were talking about that issue a few posts up.  Cards older than 1060 don't work with HSR.  You'll have to use CcminerKlaust as described in this post: https://bitcointalk.org/index.php?topic=1777336.msg32342559#msg32342559

Man, hsrminer_neoscrypt_fork works even with GTX 970, and funny little gpu like GT 1030, what are you talking about?  Smiley

The problem could be that most scripts like MrPlus' one and Nemos download and use hsrminer_neoscrypt_fork_hp.exe - it's high priority version of my fork, and it clearly states that "You will get more hashrate, but also high load on GPUs! If you have troubles with it - use normal process priority version (or specify -c 2 for this version)". And many new users don't have a clue about hardware and how to properly use it, so they just overclock their gpu to high values AND reduce PL%, and press button to start miner -  then "all of a sudden" gpu become unstable under high load and crash.

You can make changes to script so it will download normal process priority version to lower GPU's load. If you still have problems with hsrminer_neoscypt_fork, post questions to my topic, I will help you.

I didn't know of the normal process priority version.  It may be more stable.  I based my comment on other users posting of crashes.  Thanks for letting us know of the other version.

The miner itself is stable - otherwise my topic would be full of crash reports from users. Only reasons why it can crash - if GPU is overclocked way too much and can't handle high load ( because neoscrypt algo is very gpu-intensive) or user specify random intensity value with -i option (because program is optimised for certain values and usually you don't need to tweak intensity as default values are already fine-tuned for 1060,1070,1080(Ti) Gpus).

As for that user's comment - he is a newbie with 1 post and a rig of 4 x 1050 Ti and he claims that "Hsrminer is sick and buggy as hell"? It's not even funny - hundreds of users with all types of GPUs (starting from 9xx to Titan XP) are using it without any problems.

Of course only MrPlus can deside what miner he will include into his package, but I don't think he should make a decision based on fake claims of some random dude. If you guys have any issues with hsrminer_neoscrypt_fork, contact me via PM, or post questions to my topic.

It wasn't just him,  I think that the normal process priority version will work better as users push oc for the other miners and then when switching to HSR_hp it crashes.  HP is probably just fine if you optimize oc for it and not for multiple miners.. just my opinion of what is happening.

Hey Just_a_miner

Have absolutely nothing against hsrminer_neoscrypt_fork. It is a great miner and always worked well on our rigs.

The reason I changed is that not every users of NPlusMiner are advanced enough to use per algo OC using prerun and therefore get problem when they start playing with OC.

So I made a choice of stability for common users while advanced users know very well how to get hsrminer_neoscrypt_fork back in play with the adapted settings so it can play with other miners and OC with no problem.

This is just a default config choice (stable) anyone can customize the miners and I see a lot of users do so.

Just trying to avoid the kid crashing G'Ma's computer Smiley