Topic: [ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.2 (Linux / Windows) - page 38. (Read 224961 times)

@neekir

I tested MSI Armour, EVGA SC and Gigabyte OC. They work with mem +500. Sorry to hear about Zotac doing a weaker job.
Mining cryptonights is just for the purpose to teak your OC settings. The core frequencie will show a fixed number, not jumping around as with zec miners. Than you can switch to equihash.

One more thing... Maybe is something wrong with nVidia drivers. Unistall CUDA drivers (in case you installed them), unistall nVidia drivers, you can do a restart, or jump to the reinstall directly.
1.Install nVidia drivers (the last version, and no need for Geforce expirience or Phisix).
Watch in the Task manager if any nVidia Container process uses CPU more than 1-5%. Wait for it to stop (it means that all the GPU's are installed).
BTW... Idealy, you should have only one nVidia Container process runing. If there are more, when you will install CUDA in last step, all those will be replaced with only 1 instance.

2.Download Microsoft Visual studio (VC-redist) both versions x86 and x64 from www.visualstudio.com/downloads/ (the second from bottom down).
Go to "older versions" and download the 2015, the 2013 and the 2012 versions x86 and x64.
Unistall all versions of Visual Studio from Control Panel > Programs.
Install in this order: vc_redist 2012 x86 and x64, 2013 x86 and x64, 2015 x86 and x64, 2017 x86 and x64.

3.Reboot!

4.Install nVidia CUDA drivers (toolkit... whatever), last version, from here developer.nvidia.com/cuda-toolkit-archive.
CUDA drivers are needed for cryptonight miners, but is good to have them installed.
Reboot!

And PLEASE !!! FIX your PING!!! 260ms is NOT GOOD FOR MINING!!! You must have below 100ms, 30-60ms will be awsome.

Here's how to choose the best pool:
Look at pools lists like "investoon.com/mining_pools" or "miningpoollists.com", choose the top 5 pools with the most miners and hashrate (you can open each pool site and check the real stats there, at pool hashrate, not network hashrate).
Take note of the link of the stratum server for your coin on each pool (the port is not important).
Open Command prompt, and type:
ping xxxxxx.yyyyy.zzz
where xxxxxx.yyyyy.zzz is the stratum server address.
Compare the time in ms. You must choose the pool with the lowest ms value.
In this way, your shares will arrive in time to be accepted by the pool, and more coins will be gained.

Thanks a lot Onecryptoguy, I tried mem 450+ for equihash, but Zotac one does not support any where above 400, it is not stable, 500 it crashes, where 450 + runs for an hr , and it is not stable, ideally 400 is ok.

I am not mining in cryptotonight, so I did not try that settings.

Thanks for the research and sharing of cryptotonight settings.....

Sir_blacks

I understood the full logic, but temperature, I am unable to control, Let me see what additional cooling I can add, Once again thanks a lot, it has improved the performance drastically

Power 74, temp: 80 , coreclock: +160 , mem clock:+150 , fan custom curve :  65-75% auto

   GPU2  78C  Sol/s: 279.4  Sol/W: 3.60  Avg: 275.9  I/s: 153.7  Sh: 0.56   1.00 244
   GPU3  79C  Sol/s: 285.6  Sol/W: 3.58  Avg: 272.7  I/s: 152.3  Sh: 0.57   0.96 262
   GPU0  78C  Sol/s: 292.0  Sol/W: 3.59  Avg: 273.6  I/s: 152.0  Sh: 0.49   1.00 253 +
   GPU4  77C  Sol/s: 286.4  Sol/W: 3.51  Avg: 279.3  I/s: 153.7  Sh: 0.48   1.00 240 +
   GPU5  74C  Sol/s: 289.5  Sol/W: 3.44  Avg: 285.4  I/s: 155.0  Sh: 0.59   0.99 264
   GPU1  76C  Sol/s: 287.2  Sol/W: 3.47  Avg: 281.1  I/s: 154.2  Sh: 0.57   0.98 241
   GPU2  78C  Sol/s: 285.1  Sol/W: 3.60  Avg: 275.9  I/s: 153.5  Sh: 0.56   1.00 244
   GPU6  72C  Sol/s: 282.2  Sol/W: 3.27  Avg: 285.4  I/s: 152.6  Sh: 0.59   0.98 346
   ========== Sol/s: 2007.9 Sol/W: 3.50  Avg: 1953.3 I/s: 1073.3 Sh: 3.85   0.99 264

Thanks , let me tell you the  result, Once I add another cooling system.

@neekir
GTX 1060 3GB Hynix memory:
For Equihash algo, some may say "don't tweak the memory, only the core". In case of the GTX 1060 both influence the hashrate positively (tested!).
Hynix can do mem +500, even +600, but errors will start to appear, so stick with +500 as maximum OC.
The 1060 core can be set to 2050MHz with stable performance. Above that... riscky.

So... Fireup Afterburner, make this custom fan curve in settings (check Enable user defined automatic fan control): 30C-20%, 40C-35%, 50C-50%, 80C-80%, 85C-85%. Above 85% fan speed > fan will have a short life!
At GPU OC settings, Fan control > check to be lighted up the Auto button and that little wheel (user defined auto control ON).
Start XMRig to mine some ETN on a stable pool like hashvault, or nanopool, or whatever.
After starting the cryptonight miner, you will see in Afterburner the speeds of your core and memory as stable values (with ZEC miners, the core freq. varries, so you can't tune it properly).
Now, with XMRig on, in Afterburner set:
-unlink power limit and temp limit;
-power limit 85;
-temp limit 80 (or 83);
-core +100;
-mem +500.
You will see the mem has 4303 MHz.
Now you tune the core with +/-5MHz, +/-1Mhz untill you have a stable 2050MHz.
These are your maximum stable OC settings for GTX 1060 3GB.
For ex. a EVGA SC card will need somewhere between core +30 and +65 to rich 2050MHz. MSI +101/+115... etc. depending on the default clocks.
Now, you can shutdown XMRig, and start mining with DSTM's ZM some ZEC or ZCL, or whatever...
With these settings, my rig of 6 GTX 1060 3GB, in Windows 10, with ZM, stays at 300-320 sol/s on each GPU, at a ~1880 sol/s in total.
My temps are at 50-52C, but my room temp is 12-15C.

P.S.: for mining cryptonights with this GPU, set power limit 65, temp 80, core -100, mem +500. It's memory intensive, core must stay underpowered. You can have 480-490 h/s in XMRig.

i can't seem to get the intensity setting to work right.  i Have a 1080ti win7-64. It seems that when i use a setting of  .9 my hashrate will go from 700sol/s to 120sol/s.. lol which is a bit to low for %90 intensity.. Then if i use .1 which is %10 intensity i get 65sol/s. Is anyone getting this to work properly on their cards?

@dstm

can you please output information about new job work IDs from pool
for example in line
> : indicates that a new job was received

Thanks in advance.

i will quote from default cfg bring from dstm, see in your directory were you was extract zm . cfg

@neekir do not increse clock to much aka +160-170 is aceptable one for 74 Power limit you risc 2 thing: get agen more temperature & fry gpu force him to goo on higher clock with less power

make this hapen, i know easy to say hard to be done
then you will be fine, so do this:
- Power limit 74, temp limit 80, core clock +160(my udvice max 165), mem clock 0(100-150), Fan: 75% fixed
glad i was can help you
cheer

anyone has tried intensity option in start.bat in v0.6? i still not understand . CCminer for neoscrypt easy just put -i 17.6 <--(value)

I increased core clock to 190, rest of the settings same, need to see it stays with out crashing, Thanks a lot for giving me valuable knowledge

Coreclock +190 mem 0 fan 75% fixed temp 72 power 74

  ========== Sol/s: 2001.4 Sol/W: 3.43  Avg: 1974.2 I/s: 1068.0 Sh: 5.27   1.00 293
   GPU0  71C  Sol/s: 282.7  Sol/W: 3.51  Avg: 276.3  I/s: 150.0  Sh: 0.69   1.00 242
   GPU1  71C  Sol/s: 287.3  Sol/W: 3.38  Avg: 285.5  I/s: 153.9  Sh: 0.52   1.00 250
   GPU2  71C  Sol/s: 287.9  Sol/W: 3.51  Avg: 280.2  I/s: 151.5  Sh: 0.57   1.00 243
   GPU4  71C  Sol/s: 284.4  Sol/W: 3.46  Avg: 281.4  I/s: 153.2  Sh: 0.75   1.00 259
   GPU3  71C  Sol/s: 278.3  Sol/W: 3.52  Avg: 276.0  I/s: 150.9  Sh: 0.74   1.00 252
   GPU5  71C  Sol/s: 290.4  Sol/W: 3.37  Avg: 288.7  I/s: 156.0  Sh: 0.86   1.00 271
   GPU6  69C  Sol/s: 290.8  Sol/W: 3.29  Avg: 286.6  I/s: 154.3  Sh: 1.03   1.00 540
   ========== Sol/s: 2001.9 Sol/W: 3.43  Avg: 1974.7 I/s: 1069.8 Sh: 5.16   1.00 293

@neekir well thx to @promomei here your prof see how much important is temperature
@promomei perfect proff memory it not matter


to get better result you need to lower your temperature to at list ~62 best case scenario for 80 Temperature Limit & for sure that is not easy for you be in india were humidity + temp are usual high from what i know .... so now i think you get it

Power Limit it set a limit what GPU it will goo base on Temperature limit set aka that 2 value are link togheter what are translated in I/s: = intensity / sec what influence Sol/s

@neekir set this Power 74, temp 80, core clock +160 , mem clock 0 , Fan: 75% fixed = you will get slice more

cheers

Im still using 0.5.8
works great on 1060 6gbx 6 (micron)

========== Sol/s: 1913.9 Sol/W: 3.33  Avg: 1900.9 I/s: 1018.4 Sh: 5.16   1.00 72
   GPU0  42C  Sol/s: 313.6  Sol/W: 3.35  Avg: 318.2  I/s: 170.0  Sh: 0.93   1.00 78
   GPU1  42C  Sol/s: 316.0  Sol/W: 3.32  Avg: 316.3  I/s: 169.6  Sh: 0.89   0.99 78
   GPU2  45C  Sol/s: 305.2  Sol/W: 3.30  Avg: 314.9  I/s: 168.7  Sh: 0.80   1.00 78
   GPU3  42C  Sol/s: 313.0  Sol/W: 3.33  Avg: 316.9  I/s: 169.9  Sh: 0.89   1.00 63  +
   GPU4  41C  Sol/s: 315.4  Sol/W: 3.31  Avg: 315.5  I/s: 169.3  Sh: 0.87   1.00 62  ++
   GPU5  32C  Sol/s: 327.6  Sol/W: 3.35  Avg: 319.1  I/s: 171.0  Sh: 0.79   1.00 78

here is my msi burner settings

https://ibb.co/igjny7

Thank you Sir_Blacks

I figured out, my GPU's were kept in a closed box with air passage, I kept the side cover fully open, so that it is more or less equal to a open rig

Things have improved, I never got this much good result .

>  GPU4  71C  Sol/s: 270.8  Sol/W: 3.75  Avg: 251.8  I/s: 147.4  Sh: 0.53   0.97 254 +
>  GPU3  71C  Sol/s: 272.4  Sol/W: 3.72  Avg: 249.6  I/s: 145.4  Sh: 0.52   1.00 293
>  GPU1  71C  Sol/s: 286.3  Sol/W: 3.60  Avg: 263.9  I/s: 151.4  Sh: 0.42   1.00 361
>  GPU5  71C  Sol/s: 283.7  Sol/W: 3.66  Avg: 265.5  I/s: 154.2  Sh: 0.50   1.00 422
>  GPU2  71C  Sol/s: 277.0  Sol/W: 3.69  Avg: 259.5  I/s: 149.3  Sh: 0.48   1.00 322
>  GPU0  71C  Sol/s: 276.1  Sol/W: 3.75  Avg: 248.8  I/s: 146.9  Sh: 0.46   1.00 301 ++
>  GPU6  70C  Sol/s: 277.6  Sol/W: 3.53  Avg: 261.9  I/s: 152.0  Sh: 0.72   1.00 614
   ========== Sol/s: 1943.8 Sol/W: 3.67  Avg: 1801.0 I/s: 1046.6 Sh: 3.63   1.00 366

This is the current settings

Power 74, temp 72, core clock +160 , mem clock 0 , Fan: 75% fixed. Is the intensity/sec is ok, still is it very bad?  

Thanks a ton for your help

Thank you Sir_blacks and Atronoss

I am trying this now......Will let you know the result in one hr. I can see improvement with hash comparing to the old one.

I am from India, I am using Nicehash India Pool , not sure, Which pool is better

I'm also mining with Claymore on two reference AMD R9 270 2GB cards. As I said, reference design, so with very poorly designed cooling, they both running between 80-85°C and both are running fine for whole one year

So no, I think his fans will probably not die in months, nor the card. Until 80°C it should be fine, he just can't reach higher nVidia boost clock speeds.

Best way for him is to try reduce power limit and compensate with higher boost clock on core, and highest possible stable memory clock.

see @neekir there is your problem

your temp limit is 78 (you are already on limit) & you let fan auto speed aka your gpu are already to 77 so is exactly what i think aka you get low hspeed because miner it lower intensity to protect (I/s: = intensity / sec)

try this setting:

- Power Limit 74 = this it suppose to lower heat to a max ~62 my be ~63 depend how is your ambience temp
- set Fan speed to a permanent 65 - 70% (tune your value till you see you get lower temperature) let them work on auto sometime it work badly better keep them on a specific speed always coz your rig is never stop & never rest

idea is to high temperature it lower gpu & miner intensity because it try to self protect so who say keep them on high temp is good tell them to goo to hell .... yes actual gpu can run on high temp BUT that not means it is normal most bad enemies for all electronic is temp & dust
- Temperature limit 80
- Lower GPU clock to 160 that means it will lower temperature to & it will alow to increase memory probabil bigger value but i say agen on equihash speed memory count less

yes hynix is worst vs samsung from any point of view but see what i tell about equihash algoritm ... if you was mining eth i say yup is relevant but in our case nop

the point is you need to get a balance between power usage > temp > fan speed other way you get scrue up your self losing a lot from income & killing your equipment, tuning you can make it with out stop & start miner app if you using windows & msi afterburner but it possible to crash till you find right combo ...

same get a better pool with better latency 250 is bad bad bad special for a dedicate rig

hope it help

See the drop, the current settings is this\

 GPU0  73C  Sol/s: 249.1  Sol/W: 3.54  Avg: 259.6  I/s: 132.4  Sh: 0.30   1.00 253
>  GPU1  69C  Sol/s: 243.3  Sol/W: 3.54  Avg: 261.9  I/s: 132.6  Sh: 1.19   1.00 262
>  GPU2  70C  Sol/s: 247.0  Sol/W: 3.60  Avg: 260.5  I/s: 132.7  Sh: 0.90   1.00 248
   GPU3  75C  Sol/s: 249.8  Sol/W: 3.62  Avg: 255.0  I/s: 132.6  Sh: 0.90   1.00 263
   GPU4  72C  Sol/s: 236.0  Sol/W: 3.55  Avg: 254.5  I/s: 132.1  Sh: 0.30   1.00 242
   GPU5  69C  Sol/s: 247.4  Sol/W: 3.58  Avg: 261.9  I/s: 134.0  Sh: 0.60   1.00 271
   GPU6  69C  Sol/s: 247.0  Sol/W: 3.48  Avg: 261.9  I/s: 132.9  Sh: 1.19   1.00 279 ++


----

core clock: +110  Mem:+300 ,power :80 and temp :70 ,fan :auto

What is your ambient temperature anyway?

I have 3 CPU fans which pushes the air out , table fan and exhaust fan too , no AC , power will not support for an AC more

I tried all settings what I saw in youtube and all blogs in the world.

None of them are working, this is one of the worst cards I received.Zotac 1060 3GB with Single fan

I changed temperature to 70C, hashing went down to 235-250

Suggest me a better setting to work in less temperature and reasonable hashing.

WTF is with some of you, my temps don't exceed 50C, and you're running your cards at 75+?
How will it be when the outside temp is like 40C at summer?

Also your fans will die in like few months.
Do a custom fan curve for gods sake and open your windows, or install an AC for the summer months atleast.