Author

Topic: EWBF's CUDA Zcash miner - page 257. (Read 2164327 times)

member
Activity: 108
Merit: 11
December 28, 2016, 10:45:03 AM
EWBF 0.0.8


2 x GTX 1080 OC
GPU0: 450 Sol/s GPU1: 451 Sol/s
Total speed: 901 Sol/s


Thanks a lot ! Wink
legendary
Activity: 1108
Merit: 1005
December 28, 2016, 10:39:14 AM
I switched to this miner and got 350sols with a Gigab G1 1070.. (no OC, sadly 100% tdp, i cant lower that without trashing speeds)

BUT... i left the miner overnight and i found it crashed two hours later... I lost the whole night of mining. My fault for testing it unattended.


No probably not your fault, try running more instances (2-3), you will get speed increase and if one instance crashes than you still have the other.

It does not help, because whole machine is crashing, not just miner.
So, improved stability in next release, would be much appreciated.
full member
Activity: 198
Merit: 160
December 28, 2016, 10:12:47 AM
Hello, I am trying to mine with a GTX 950 (obviously not trying to make any money Tongue). When I try running the miner I get 'ERROR: Cannot run workers'. I am on a 64-bit Windows 7 PC and I am trying to mine on AntPool.

Thanks for any help

It's probably problem with gpu drivers, try update him or reinstall.
sr. member
Activity: 463
Merit: 250
December 28, 2016, 09:31:59 AM
I switched to this miner and got 350sols with a Gigab G1 1070.. (no OC, sadly 100% tdp, i cant lower that without trashing speeds)

BUT... i left the miner overnight and i found it crashed two hours later... I lost the whole night of mining. My fault for testing it unattended.


No probably not your fault, try running more instances (2-3), you will get speed increase and if one instance crashes than you still have the other.
hero member
Activity: 968
Merit: 624
Still a manic miner
December 28, 2016, 07:30:30 AM
I switched to this miner and got 350sols with a Gigab G1 1070.. (no OC, sadly 100% tdp, i cant lower that without trashing speeds)

BUT... i left the miner overnight and i found it crashed two hours later... I lost the whole night of mining. My fault for testing it unattended.

sr. member
Activity: 455
Merit: 250
December 28, 2016, 12:12:05 AM
Hello, I am trying to mine with a GTX 950 (obviously not trying to make any money Tongue). When I try running the miner I get 'ERROR: Cannot run workers'. I am on a 64-bit Windows 7 PC and I am trying to mine on AntPool.

Thanks for any help
what you trying to mine on antpool?it is bitcoin and litecoin pool mostly.950 for altcoins mostly.don't even try antpool without ASIC.
awww.they added zcash...then you need to add worker in WORKERS.
jr. member
Activity: 61
Merit: 1
December 27, 2016, 11:47:38 PM
Hello, I am trying to mine with a GTX 950 (obviously not trying to make any money Tongue). When I try running the miner I get 'ERROR: Cannot run workers'. I am on a 64-bit Windows 7 PC and I am trying to mine on AntPool.

Thanks for any help
sr. member
Activity: 455
Merit: 250
December 27, 2016, 11:47:20 PM
something like that...




Power consumption is too much. I can get 375 sols with 50 % tdp. So you get 35 sols more  with extra 35 % tdp.
yeah but i use multipoolminer which mine all the algos on miningpoolhub, so i need 100 tdp for changing algos and miners without crashing.also - my electricity is only 0.05$/kWt sooo...it's not a problem=)
full member
Activity: 186
Merit: 100
December 27, 2016, 12:44:41 PM
something like that...




Power consumption is too much. I can get 375 sols with 50 % tdp. So you get 35 sols more  with extra 35 % tdp.
sr. member
Activity: 455
Merit: 250
December 27, 2016, 09:51:36 AM
something like that...



legendary
Activity: 1764
Merit: 1024
December 27, 2016, 09:30:49 AM
Compared to EQM, .7 uses ~4.5% less power and produces 4% less hash. So roughly the same give or take, slightly less efficient. This was done running two instances of EWBF.

Right now Nicehash is paying 8% less then mining direct. This miner has a 2% fee, so this is almost a complete wash. Really comes down to what you rather do, mine coins or get paid bitcoin direct. The net result is currently the same, but that can fluctuate depending on if and when you sell your coins.

I haven't had any crashing so far, I'm guessing that's due to people messing with TDP settings and strangling their GPU (it doesn't have enough power to keep the memory OCs stable). Memory OCs do not handle less power nearly as well as GPU OCs. You can also have crashing with stock clocks and lowering your TDP to a bare minimum (manufacturers don't rigorously test settings like this so their 'stock' OCs may not work properly).
sr. member
Activity: 574
Merit: 250
Fighting mob law and inquisition in this forum
December 27, 2016, 08:09:56 AM
ioglnx Which driver do you use? i'm using 376.33.

I'm using th elatest hotfix driver which has the workaround for Folding@Home before that I used 376.09 because 736.33 was again based on 375_rev instead 376_rev like he 09.
I will try to restore my snapshot with old driver I also think this driver is broken since it makes no sense on all these errors I see.
full member
Activity: 243
Merit: 105
December 27, 2016, 07:42:51 AM
Hi everybody!

A couple of answers to your questions:
Expected speed ~360 on stock 1070 ~250 on stock 1060
Stability, miner now work with "intensity" it better utilize gpu and as result it can make strong overclock less stable.
DevFee, miner works for devfee 2% of time. But difficulty on devfee pool can be significantly lower or higher than on your pool.

It would be good, if would be intensity param, to fine tune cards.
member
Activity: 108
Merit: 11
December 27, 2016, 07:38:40 AM
Quote
CUDA: Device: 3 Thread exited with code: 77
CUDA: Device: 4 Thread exited with code: 77
CUDA: Device: 2 Thread exited with code: 77
CUDA: Device: 1 Thread exited with code: 77
CUDA: Device: 0 Thread exited with code: 77
ERROR: Some workers are stopped. Attempt to restart: 1
CUDA: Device: 0 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 1 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 2 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 3 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 4 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 0 Thread exited with code: 46
CUDA: Device: 1 Thread exited with code: 46
CUDA: Device: 2 Thread exited with code: 46
CUDA: Device: 3 Thread exited with code: 46
CUDA: Device: 4 Thread exited with code: 46
ERROR: Some workers are stopped. Attempt to restart: 2
CUDA: Device: 0 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 0 Thread exited with code: 46
CUDA: Device: 1 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 1 Thread exited with code: 46
CUDA: Device: 2 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 2 Thread exited with code: 46
CUDA: Device: 3 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 3 Thread exited with code: 46
CUDA: Device: 4 GeForce GTX 1070, 8192 MB M0
CUDA: Device: 4 Thread exited with code: 46


to high overclocked !!
full member
Activity: 198
Merit: 160
December 27, 2016, 07:30:01 AM
Hi everybody!

A couple of answers to your questions:
Expected speed ~360 on stock 1070 ~250 on stock 1060
Stability, miner now work with "intensity" it better utilize gpu and as result it can make strong overclock less stable.
DevFee, miner works for devfee 2% of time. But difficulty on devfee pool can be significantly lower or higher than on your pool.
legendary
Activity: 1106
Merit: 1014
December 27, 2016, 07:05:51 AM
ioglnx Which driver do you use? i'm using 376.33.
sr. member
Activity: 574
Merit: 250
Fighting mob law and inquisition in this forum
December 27, 2016, 06:58:40 AM
Nonsense answer  Roll Eyes and Linux easily gets a kernel panic. So it would be the same result in the end!
I don't want to switch my OS because I ran stable witt 0.0.6b
Your cards are most likely unstable. You wrote earlier that you're running them at +150/+650. It's waaaay too much for most cards with Micron chips. Zcash miners, especially EQM miner, don't stress the cards very much, so many users were able to get away with very high overclocks. But those aren't rock solid stable overclocks, not even close. If they'd test those OC clocks in a dozen of games they'd have crashes in half of them. Same with mining, some algos/miners are more stressful and some are less. I've had this 0.0.7 version running for ~12 hours non stop with my 1070s and no issues whatsoever. That's because my OC is thoroughly tested.

What I don't like about this miner is that there might be something wrong with devfee.



Not sure how it works but it's like it's taking too much. In only 15 minutes running 0.0.7 pointed at Dwarf pool, I've got 35 shares accepted by the Dwarf, and 6 Devfee shares accepted. That's not 2% fee. But this might have something to do with difficulty maybe? Higher diff at Dwarf pool and lower diff at the dev's pool? Anyway, would like to see an explanation to this.
If you read more carefully you would have noticed that i made other runs with no overclock - I repeat it also happened with no OC! and 0.0.6 runs stable even two instances.
legendary
Activity: 3248
Merit: 1070
December 27, 2016, 06:03:59 AM
Thanks dev Smiley .

The best setting for nvidia gtx 1070 with version 7 ? .

Now i have gpu clock +150 , mem clock +700 , Power Limit 55% for 360 sol .

With these settings I'm doing 300-310 sols. What am I doing wrong???
5 x evga gtx 1070 sc
h81 pro btc
celeron 1820
windows 10 64
nvidia driver 376.19

tks!!!


it's the model, the number of pin your card have and the memory type, if you have bad memory like micron and not samsung your memory oc will not be as a strong
legendary
Activity: 1106
Merit: 1014
December 27, 2016, 05:52:20 AM
Nonsense answer  Roll Eyes and Linux easily gets a kernel panic. So it would be the same result in the end!
I don't want to switch my OS because I ran stable witt 0.0.6b
Your cards are most likely unstable. You wrote earlier that you're running them at +150/+650. It's waaaay too much for most cards with Micron chips. Zcash miners, especially EQM miner, don't stress the cards very much, so many users were able to get away with very high overclocks. But those aren't rock solid stable overclocks, not even close. If they'd test those OC clocks in a dozen of games they'd have crashes in half of them. Same with mining, some algos/miners are more stressful and some are less. I've had this 0.0.7 version running for ~12 hours non stop with my 1070s and no issues whatsoever. That's because my OC is thoroughly tested.

What I don't like about this miner is that there might be something wrong with devfee.



Not sure how it works but it's like it's taking too much. In only 15 minutes running 0.0.7 pointed at Dwarf pool, I've got 35 shares accepted by the Dwarf, and 6 Devfee shares accepted. That's not 2% fee. But this might have something to do with difficulty maybe? Higher diff at Dwarf pool and lower diff at the dev's pool? Anyway, would like to see an explanation to this.
full member
Activity: 243
Merit: 105
December 27, 2016, 05:36:10 AM
The crashes occurred randomly and without connection to current clock speeds.
I had this night 10 blue screen, 38 crashes..some times directly at start, sometimes 5min after and then after 2h sometimes.

Profit for me due to these crashes and bluescreen halved ..switching back to 0.0.6b meanwhile
Use linux and you won't get bluescreens lol

Nonsense answer  Roll Eyes and Linux easily gets a kernel panic. So it would be the same result in the end!
I don't want to switch my OS because I ran stable witt 0.0.6b

I think that it is windows only issue, or hardware problems( sometimes 1070 can drain up to 50watt from pci-e ), all my linux rigs with overclock run well with 0.0.7
Jump to: