Author

Topic: EWBF's CUDA Zcash miner - page 130. (Read 2164327 times)

newbie
Activity: 25
Merit: 0
June 16, 2017, 11:26:34 AM
Ok, I think I am ready to start tinkering with my 1080ti's

What sort of upper limit results have people seen with the 1080ti? Power and cooling is not an issue, so I'd like to push the cards as much as I can without destroying them.  

So far I'm seeing +/- 1430 sol/s @ 49C -- stock settings.

2x Aorus GTX 1080ti Xtreme edition
8GB DDR4 Kingston HyperX
ASRock Z270 K6 Gaming
2.9 GHz Celery processor
newbie
Activity: 7
Merit: 0
June 16, 2017, 09:40:41 AM
1070 performance on 0.3.4b? Anyone?  Cheesy

I just received my first 1070 yesterday.
(an EVGA FTW, just running at stock speeds now)
Still new at this, have been using EWBF for about a week.

EWBF says  ~445 Sol/s, 150W.  The watts figure closely matches my Kill-A-Watt. The Sol/s seem reasonable.

Good luck!

i have 3 EVGA 1070, do like on screenshot and you get 500 sol ) On Samsung Memory ~505, All my card have Micron (


hi bonosurgut, thanks for the info.
I just recently starting playing with card settings again.
I saw someone else posted about using:
   power 81
   core clock +200
   memory +600
   fan 80%
I'm using that on both 1050ti and 1060 with good initial results.

However, in this new rig with 3 cards (1050ti, 1060 and 1070), when I use Afterburner to change settings on the 1070, the system becomes unstable. The changes to 1050ti & 1060 work without issue, but even minor changes to 1070 affect Windows - it locks up for several seconds, EWBF gives error messages, tries to restart the GPUs, work goes to 0.  I've only given a minute or so to recover, then usually reboot.

Possibly I need a different version of Afterburner (currently 4.30) or Nvidia drivers (382.05).

This is on Win10EDU 64bit. Was working well for a week before adding the 1070. Power supply is 750W, should be fine.  I'm working toward getting identical cards in each system, but it will be a while. (I thought maybe the extra complexity of different hardware leads to untested situations)

I see that your settings go in a different direction, will research this further. Thanks for any suggestions.
newbie
Activity: 14
Merit: 0
June 16, 2017, 08:36:36 AM
Hi, could somebody tell me what's solver # meaning? differents mining alg?

solver 0 = most recent GPU

Thx, could you be more specific? how NOT so recent should be my gpu to use 1 or 2 or 3? does it have to do with cuda CC ver? or just gpu type? is there any table of usage?
newbie
Activity: 2
Merit: 0
June 16, 2017, 08:25:53 AM
I don't understand how I use this Program to mine other Coins than ZEC? For instance I want to try and mine HUSH but I don't see any information on how I would switch coins. Any help would be appreciated!

I just changed the pool address and port in the .bat file and I've been mining ZEN without any problems.
newbie
Activity: 4
Merit: 0
June 16, 2017, 08:12:35 AM
I don't understand how I use this Program to mine other Coins than ZEC? For instance I want to try and mine HUSH but I don't see any information on how I would switch coins. Any help would be appreciated!
YIz
hero member
Activity: 686
Merit: 502
June 16, 2017, 07:11:53 AM
Has anyone tried this miner on a GTX 780 TI?
sr. member
Activity: 882
Merit: 310
June 16, 2017, 06:56:56 AM
Every card asic vary thus power draw and voltage and clocks (boost 3.0) vary.
So it can vary clock and power draw, as I have seen on my cards, one card can have 1900 while second almost 2000mhz, with same settings.
full member
Activity: 430
Merit: 100
June 16, 2017, 06:40:56 AM
Hey guys, i got 2x 1070 and i'm getting some lower results than all of you

i had to put power on 90% +130 clock +680 memory in order to get 435sol

can anyone share how they managed to get the best results for a 1070? thanks

power 85
core clock 120
memory 700
fan 80%

getting 965 sol/s with 150W -> 3.25 SOL/W

or

power 61
core clock 150
mem 400
fan 70%

getting 855 sol/s with 108W -> 3.9 SOL/W

or

power 58
core clock 150
mem 400
fan 50%

getting 825 sol/s with 101W -> 3.9 SOL/W

it just depends how do you have to pay per KW/Hour ... just try your cards, and after every change you made, wait about 10-15 mins to get a accurate data from the miner. I also let the miner restart every 30 Mins and take care, that my cards are still under 60C, because i still use them for gaming.

Using these:
power 85
core clock 120
memory 700
fan 80%

and getting 850 sol/s Sad
i got an EVGA and Palit cards, they perform very good on ETH mining so i find this odd.
what else can i try? i'm using windows btw.

also, is this miner legit? does mh/s stand for sol/s? Thanks! https://www.cryptocompare.com/mining/calculator/zec?HashingPower=2000&HashingUnit=H%2Fs&PowerConsumption=0&CostPerkWh=0.12
sr. member
Activity: 882
Merit: 310
June 16, 2017, 06:22:38 AM
I have 2x GTX 1060 3GB (for now, couldn't plug 2 more), and 285-290 sol per card, with 1930mhz and +400mhz mem, +10 power limit.
They are using 115-118W, so it's quite a lot, and I wouldn't bother, but I have only 700W power supply with 650W on 12V, so I want to reduce power consumption. Should I lower power limit and clocks? Or only power limit ? Now voltage is around 1.03-1.04V.
I want to have about 80-90W/per card so total cards has around 360W and system around 450-460W (not to stress out power supply to the max), and to not reduce the speed to much, is it possible ?
Or better go ETH mining/dual mining with nvidia ? With AMD it was a good option, but I don't know if it will be a good option with NV.
Also is --fee option still work, or it was only working in previous versions?
newbie
Activity: 6
Merit: 0
June 16, 2017, 05:58:59 AM
Hey guys, i got 2x 1070 and i'm getting some lower results than all of you

i had to put power on 90% +130 clock +680 memory in order to get 435sol

can anyone share how they managed to get the best results for a 1070? thanks

power 85
core clock 120
memory 700
fan 80%

getting 965 sol/s with 150W -> 3.25 SOL/W

or

power 61
core clock 150
mem 400
fan 70%

getting 855 sol/s with 108W -> 3.9 SOL/W

or

power 58
core clock 150
mem 400
fan 50%

getting 825 sol/s with 101W -> 3.9 SOL/W

it just depends how do you have to pay per KW/Hour ... just try your cards, and after every change you made, wait about 10-15 mins to get a accurate data from the miner. I also let the miner restart every 30 Mins and take care, that my cards are still under 60C, because i still use them for gaming.
newbie
Activity: 6
Merit: 0
June 16, 2017, 05:47:47 AM
Hi, question. Having a 1070er Rig (6Cards). In the mining software the cards are reporting stable 475 H/s per Card. But in the pool (flypool) are big drops and highs, its not a stable line like in the mining software.

Any ideas why? I see the CPU is under full throttle (Intel Pentium). Is here a i3 better for cuda mining? Also having only 4GB DDR3 on Board. Maybe 8GB better?

dont worry Smiley everything is fine! its a issue due to flypool, just wait a minimum of 24h nonstop mining to get a better average.

You better use the flypool android app, the app hat more accurate date, because it uses the flypool api.

475sols is very ok, you should take a look to the power usage W/Sols .... go down with power limit 85, overclock gpu about 120, ram about 700 set your fans to 80%
full member
Activity: 430
Merit: 100
June 16, 2017, 05:35:02 AM
Hey guys, i got 2x 1070 and i'm getting some lower results than all of you

i had to put power on 90% +130 clock +680 memory in order to get 435sol

can anyone share how they managed to get the best results for a 1070? thanks
newbie
Activity: 2
Merit: 0
June 16, 2017, 05:29:32 AM
Hello guys. A the begining I'm sorry for my english.
I have a problem with my mining rig
My conf:
6x1070 MSI ARMOR GPU
Celeron
4GB DDR4
120SSD
MSI Z170 MAINBOARD
cooler master 1200w platinum
W10

With 6 cards my miner crash about 10 minutes after start, with 5 cards miner works good (18h-24h) without crash. When miner stop one graphic from msi afterburner are dissapear, then i must restart computer.

WITH 5 CARDS:
http://imgur.com/a/Sfzdx
MSI AFTERBURNER:
http://imgur.com/a/B1NN1
CRASH:
http://imgur.com/a/dSOYu
MY EWBFS:
http://imgur.com/a/Uc7iH



Maybe I have a bad conf in mainboard bios?
Im using newest gpu drivers.
Do you have any tips what should i do to work on 6 cards ?
MY MINER: http://imgur.com/a/yKemv

THX for help good people


full member
Activity: 240
Merit: 100
June 16, 2017, 05:18:49 AM
Is there a way to fix gpu not restarting properly, it just keeps saying its restarted but thread error 44 and not mining again. I'm now starting all gpus separately so if only one fails the rest are still mining. That taskill loop/restart workaround doesn't want to work.

To have 480sols on 1070 you need to OC, mine is +130/700 and 76 power, voltage is ~0.943, gpu clock is around 1950Mhz, memory 2250 in gpuz.
700 but it crashes. better have only 450s/s and no crash. What do yo think?
full member
Activity: 336
Merit: 100
June 16, 2017, 03:17:29 AM
MSI GTX 1070 Gaming X 8g > 470-480 Sol/s @ 160W (2.94 Sol/W) - doesn't look to bad i think Smiley

tweaked with afterburner ? Cant be stock, screenshot of settings please ?
member
Activity: 126
Merit: 10
June 16, 2017, 02:37:29 AM
Hi, question. Having a 1070er Rig (6Cards). In the mining software the cards are reporting stable 475 H/s per Card. But in the pool (flypool) are big drops and highs, its not a stable line like in the mining software.

Any ideas why? I see the CPU is under full throttle (Intel Pentium). Is here a i3 better for cuda mining? Also having only 4GB DDR3 on Board. Maybe 8GB better?
member
Activity: 126
Merit: 11
June 16, 2017, 12:42:14 AM
How sharing is more than 100% possible?  Angry


They are just estimating how many shares it would take to find the solution based on the current difficulty by some statistics.  That doesn't mean that you will find the "Solution" always in that number of shares.  Some times the pool is lucky and you find the "Solution" in only 800,000 shares, other times its not lucky and takes 23,300,000 shares.
newbie
Activity: 1
Merit: 0
June 15, 2017, 11:49:14 PM
How sharing is more than 100% possible?  Angry
https://www.dropbox.com/s/xc1ag050341f8xh/shared.png
newbie
Activity: 6
Merit: 0
June 15, 2017, 10:44:29 PM
Hi, I've got a few questions for anybody who uses the miner on linux/ubuntu, don't have very much experience at all with using linux. I've got a few 1080ti's, and I've installed the latest proprietary Nvidia drivers. I've seen people mention installing cuda as well though. Is that necessary, and if so, does it come along with the regular nvidia drivers, or is it something extra to install? Also, compared to the windows version of the miner, do all the same commands work? As in things like fee, pec, etc. Just add them to the config file as extra parameters, under the common section? In addition, is there some convenient place to see the stats as they come in, like the windows command prompt shows you when it's running on windows? I know you can go and see the web console, and see a snapshot of the stats, but there's no equivalent of the constantly running stats that the windows version had?

If you are thinking about mining on linux, you should check out this distro specifically made for nvidia gpu mining https://bitcointalksearch.org/topic/os-nvoc-easy-to-use-linux-nvidia-mining-1854250
It's the easiest one to use.  Smiley

I might look into that, saw it before, but thought I'd try it on my own first.

On that note, again, to anyone using linux/ubuntu, especially if you have 1080ti's. Does your power tend to bounce around a lot? On windows, when I ran it with a 1080, it would push right up to where I set the limit, with occasional dips. Now on ubuntu with a 1080ti, I'm seeing random drops in power of 25-50+ watts, before it goes back to around what I set it at. Probably the reason I see my clock speed jumping up and down as well.
newbie
Activity: 2
Merit: 0
June 15, 2017, 10:09:04 PM
Just today I started consistently getting this error:

Error: No devices selected

I've restarted, and removed the line --cuda_devices, and it worked once after running it, then after closing and attempting to re-run the batch file, it gives me the same error and closes out. Any ideas?

Also, before this issue, every single day the program would hang when the devfee started. It would somehow timeout and not reconnect to my pool.
Jump to: