Topic: [ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.2 (Linux / Windows) - page 105. (Read 224961 times)

This might be connected to the recent windows update - reports clearly started to increase after the update. I'm still not able to reproduce this on my systems - however I'm running constantly stability test in order to reproduce this.

Thx for reporting this. Efficiency test are surely useful for other users.
Concerning your power limit observations: If it's really stable on an PL of 50% then your stability issues might be the result of unstable power supply.

Hi Dstm,

First, thanks for your miner, the last update seemed to be pretty stable for me (Linux, 5x1070 FTW2 + 1x1070ti FTW2), but it crashed this morning with this message:

2017-11-23 06:20:35|#  GPU3  server set difficulty to: 000f0f0f0f00000000000000...
2017-11-23 06:20:35|protocol version 00000000 not supported

I was connected on Nicehash with Equihash protocol.  I was running it for 4 days staight without any problem...  Any idea what could be the cause?

Here is the command I used:

./zmminer --server equihash.usa.nicehash.com --port 3357 --telemetry \
          --user MY_NICEHASH_WALLET --pass x \
           --time >> "$log" 2>&1

Thanks!

Thank you for the help!

Add this at the beginning of your bat file:

I see, you are right. But tell me one thing, doesn't the hash rate or something decrease when you disable the dev fee? I know claymore has this feature built-in with his miners too but i've always stayed away from disabling the dev fee because he explicitly says mining will be slower. Don't know maybe the EWBF developer is not just as vocal as claymore?

Anyway yeah, this is a big factor, especially for people who have larger mining farms. I hope DSTM lowers his dev fee or adds the option to disable it completely because i believe his miner is better long term, especially in terms of stability.

Dstm, i must say great job on the miner, i really like it. Runs stable for me for days now.
I have one issue and i want to ask for a help. It happens when the power goes out and then comes back in.
My rig are configured automatically to start back up and I've put the miner in the startup folder.
The problem is the router is slower to connect to internet and the miner sees that there is no internet and just quits.
How can i configure it to start like 2 minutes later?

Thanks for the suggestion.
I agree, but I found that increased memory speed only starts to make a significant difference when used with higher core frequencies (say, 1850+ ish).
I was aiming for good efficiency which meant keeping core freq relatively low - usually ending up somewhere around 1670-1800MHz (at 800mV) depending on the quality of the chip.
Pascal seems to be very efficient at this voltage and capable of a significant overclock which is ofc desirable.
So, at these core frequencies, the reduced memory freq further reduced the total power consumption without significantly hindering the hash rate.

All of this said, it makes almost no difference anyway. You can vary a cards power consumption (reported) by up to about 4-5W though memory clock speed adjustments in my experience.

Interestingly, ZM uses less power than EWBF -and- gets a higher hash rate at the same time. It made my sols/w epeen shoot friggen rainbows.

in fact you are loosing some sols, not earning them
do you aware that ewbf miner can be run without fee?
that means  that
My 1080ti on EWBF: 660-670 sols/s  - actually is 660-670 sol/s
My 1080ti on DSTM: 670-690 sol/s - actually is 656.5 - 676 sol/s
 

Hello

Using this miner
getting 1640 sols (5x1060)

How i see its founds shares ? those +++ means that ?

on ewbf its green XD

TY

Tried DSTM's based on a friend's suggestion and boom! Slight hash rate increase.

My 1080ti on EWBF: 660-670 sols/s
My 1080ti on DSTM: 670-690 sol/s

My 1070 on EWBF: 400-405 sol/s
My 1070 on DSTM: 410-415 sol/s

Will get back to this post after a few days of testing to report stability.

Thank you for this DSTM. Well earned 2% dev fee (we would be happy if you lowered it to 1% and really blow EWBF out of the water).
#VouchingForThisMiner

Hello all.

What happens when "contime" reset but uptime continue count?

Contime means time from last disconnect from pool server?

When appeared "error recv failed: 10054 " ? Due overclocking ?

anyone has best low power settings here for 1080 , ti and 1070 ? my 1070 crashes with 70% and hot at 100% so i need a decent point.

I just tried this miner

gtx 1080
Base Clock o/s : +175
Mem Clock o/s : 0
Power limit :56%

gtx 970
Base Clock o/s : +150
Mem Clock o/s : -501 (had p0 state)
Power limit :65%

using nvidia inpector.
currently still playing around with setting to become more efficient


  GPU0  67C  Sol/s: 472.8  Sol/W: 4.27  Avg: 473.3  I/s: 252.6  Sh: 12.00 1.00 362 +
  GPU1  66C  Sol/s: 264.2  Sol/W: 2.25  Avg: 257.8  I/s: 137.8  Sh: 5.47  1.00 358 ++
   =======Sol/s: 737.0  Sol/W: 3.26  Avg: 731.1  I/s: 390.3  Sh: 17.47 1.00 360

I'm getting about 30-40 Sol/s ~extra comparing with the other miner.

Yup getting this error as well.

Same here on latest version zm 0.5.5 on
Win 10 Pro
8 x EVGA 1070 no overclock with 50% power limit
Nvidia driver v. 388.13

2 hours of work aaaaaand.....

v 0.5.5 
2017-11-22 15:21:20|gpu_id 6 1 0000000578410600 unspecified launch failure
2017-11-22 15:21:20|gpu 6 unresponsive - check overclocking
2017-11-22 15:21:20|cudaMemcpy 1 failed
2017-11-22 15:21:20|gpu_id 3 2 00000005367C3520 unspecified launch failure
2017-11-22 15:21:20|gpu 3 unresponsive - check overclocking
2017-11-22 15:21:20|cudaMemcpy 1 failed
2017-11-22 15:21:20|gpu_id 5 3 00000005C78B0200 unspecified launch failure
2017-11-22 15:21:20|gpu_id 7 4 00000005C75D9560 unspecified launch failure
2017-11-22 15:21:20|gpu 7 unresponsive - check overclocking
2017-11-22 15:21:20|cudaMemcpy 1 failed
2017-11-22 15:21:20|gpu 5 unresponsive - check overclocking
2017-11-22 15:21:20|cudaMemcpy 1 failed


Run v 0.5.4 , its oK! But ~50 sol less

thank you for your report, I will check with PL levels of course, but I doubt toy expirince will help me, cause I'm using @ 85 PL and gets like 760-770 sols from 1080ti.. and because the electricity is cheap, I'm ok if it eats 200-230 W

try add memory, not minus -50? make +300 instead
you will see you will get more sol/s
cause equihash algo loves memory speed

Hey dtsm,

I took one of the rig that was in trouble (6x1070 KFA2 EX) and went on with lengthy test, to see why it was crashing DTSM this fast even without OC on this specific brand of card, i found the solution and it's something i never had to do with any other miner, but it seems yours work in a very different way than usual so the approach needed to be unusual too.

The solution was to play with the power levels allowed on the card, i was using 56% TDP (about 112W) and OC +210/530 before on this brand of cards and it was working well with any miner as efficiency goes, i was about 3.9sols/W with EWBF (2600sols/s), after alot of playing with this i found out i had to put the cards all the way down to 50% TDP (100W) and OC +180/450 cranking up the efficiency to 4.25sols/W (2540sols/s) so the amount of sol is lower than with EWBF but the electricity cost factored in, make it 8% more efficient in term of profit, so a good 5-6% more counting your fee and the rest.

The miner (zm 5.5) now has been stable for half a day across all the KFA2 rigs, and i mostly caught up in efficiency on my 60x1070 KFA2 (~4.25s/W) vs 60x1070 EVGA (~4.3s/W), something i wasn't able to do before with EWBF (the KFA2 were 5% behind the EVGA in term of efficiency but being alot cheaper it all balanced it out )

So for me problem solved, needed the extra work, but well worth it as your fee, also learned quite a bit in the process, keep up the good work mate.

Cheers