EWBF's Cuda Equihash Miner 0.6

[B00gieman]

I cannot use This SW. Im getting always that error of cuda devices busy.

[1696295801]

1696295801

[1696295801]

1696295801

[1696295801]

1696295801

[humanv]

Equihash Miner 0.4 is fast about 10% but very unstable, need catch all bugz.

[Bonito96]

Equihash Miner 0.4 is fast about 10% but very unstable, need catch all bugz.

I agree. I back to 0.3 again.

[longphiphi]

I use --intensity 20 but it doesnot have effectively . Sol is same i:20 = i:64
Can you fix it ?

EWBF Cuda Equihash Miner v0.4

Greetings!
This is the initial release of a new miner for Equihash algorithm.
It is based on my last Zcash miner and almost completely repeats it.
Now closer to the topic:
*  Parameter "solver" is disabled.
*  Solver for equihash 200,9 has been deleted.
*  Solvers for Equihash 192,7, 144,5, 96,5 and 210,9 has been added
*  Added timestamps in the output to the screen and to the log file.
*  Argument --algo is added, allow you to select Equihash variant allowed values 144_5, 192_7, 96_5, zhash, aion. This parameter is mandatory.
*  Argument --pers is added, allow you to set personalization string (8 charachters) for equihash algorithm. This parameter is optional. Equihash by default uses the string "ZcashPoW", many other coins also use this string, but not all, some coins use their own string, and these are all differences, so for some coins, you need to set the --pers option to the required string, for example, Zero coin: https://bitcointalk.org/index.php?topic=3310714  uses the string "ZERO_PoW" and for mining this coin you need to pass arguments --algo 192_7 --pers ZERO_PoW ....
Fee is presented and by default 2%, argument --fee is presented too.

v0.4
*  Added new solver for Equihash 210,9. Available only with --algo aion (this solver is in beta status)
*  Failover pools can now be specified via command-line arguments, just repeat the credentials set, but do not shuffle them (see Example below)
*  Improved performance for algo 144_5 and maxwell gpus
*  Fixed launching of devices

v0.3
*  Added new solver for Equihash 96,5 --algo 96_5.
*  Added --color argument, allow you to disable colored console: --color 0.
*  192_7 and 144_5 solvers 0 - 10% faster, depends on the gpu.
*  192_7 require now 2.5GB vram, it should fit in 3Gb gpu under win 10.
*  144_5 require now 1.99Gb of vram.

Known issues (mostly will be fixed)
*  Equihash 96_5 Can be slower in cuda 8 versions.
*  Performance counters may display an incorrect speed during reconnection.

[HFG5521]

I'm also having the various "illegal memory, cuda devices busy, gpu restarting" errors everyone else is reporting with 0.4. Occasionally would get them with 0.3 as well but after 4-5 days. Now with 0.4 its guaranteed to happen within about a 26-27 hour period. Had the misfortune of not catching this on one rig for hours and losing precious uptime because of it. Luckily no GPU's were harmed.

[EWBF_]

I'm also having the various "illegal memory, cuda devices busy, gpu restarting" errors everyone else is reporting with 0.4. Occasionally would get them with 0.3 as well but after 4-5 days. Now with 0.4 its guaranteed to happen within about a 26-27 hour period. Had the misfortune of not catching this on one rig for hours and losing precious uptime because of it. Luckily no GPU's were harmed.

This is the new solvers, and I change them, in search of a way to make them better. So yes, every new version can be more or less stable, sorry for it guys. I continue this work until a satisfactory result is achieved.
On the intensity, it does not work right now, due to many solvers it will be completely changed and applied in the next version.
For everyone: I read this topic and see the errors that you posting here, so thanks for that, it really helps.

[CuChO]

Hello back here giving g a quick feedback :
0.3 and 0.4 works like a charm now using 144_5 algorithm  doesn't matter which coin
Super stable I get 123 sols in 3 gtx1080 each one has different core clock, evga SC +130 core, Gigabyte gaming g1 +140 core and asus strix +150 core all three
running @650 memory. Not rejected shares at all.

Now with 0.4 I switch to different coin/algorithm  192_7
I got a couple rejected  shares around 10 on the first day , now this algo won't accept  the clocks I was using,
Same clocks will cause to exit threat and just give continuously  errors.
I lower clocks to +100 core and memory to 550 in all cards, second day works with out error, rejected  shares still keep increasing are now @26
I get solid 21 sols per gxt1080

I'm using windows 10 profesional
I5 6400
16gb ddr4
500gb ssd
Asus motherboard

"""Sorry if I misspell something, I mainly speak Spanish"""

[Bonito96]

@EWBF_ 
Next version, Do you have for cuda 9.2 ?

[Zheko87]

How many sol/s on 144_5 algorithm can I squeeze with 1060 in overclocking?

[Regulater]

Hello, since version 0.4 I'm having problems with stability.

Code:

18.07.2018 09:57:52 Temp: GPU0 54C GPU1 55C
18.07.2018 09:57:52 GPU0: 48 Sol/s GPU1: 27 Sol/s
18.07.2018 09:57:52 Total speed: 75 Sol/s
18.07.2018 09:57:57 CUDA: Device: 1 Thread exited with message: "an illegal memory access was encountered"
18.07.2018 09:57:57 CUDA: Device: 0 Thread exited with message: "an illegal memory access was encountered"
18.07.2018 09:57:57 WARNING: Looks like GPU0 are stopped. Restarting...
18.07.2018 09:57:57 INFO: GPU0 are restarted.
18.07.2018 09:57:58 WARNING: Looks like GPU1 are stopped. Restarting...
18.07.2018 09:57:58 INFO: GPU1 are restarted.
18.07.2018 09:57:58 CUDA: Device: 0 Thread exited with message: "all CUDA-capable devices are busy or unavailable"
18.07.2018 09:57:58 CUDA: Device: 1 Thread exited with message: "all CUDA-capable devices are busy or unavailable"
18.07.2018 09:57:59 WARNING: Looks like GPU0 are stopped. Restarting...

After this message autorestarting loops until manual restart.

I'm using same OC settings as with v0.3 which were rock solid (up to 16 days of continous mining)

ATM I lowered OC a bit and testing if problem still occurs.

I have this problem, with 0.3 work full time, 0.4 same OC and have problems.

Have same problem wtih version 0.4. I didint have that problem with 0.3 with same OC. after a while, a problem appeared...

Exact same issue here. Updating to .4 caused instability with the OC. Had to lower the OC across my rigs that were running stable multiple day runs on .3.

[NiklasFalk]

How many sol/s on 144_5 algorithm can I squeeze with 1060 in overclocking?

I got 203  Sol/s on 9x1060 #GB with micron memory, 22.5 average (140 core, 1400 memory and 80W settings in simplemining.net). Add 10% for 6GB I guess. But ETH still pays better than BTG for me.

[dushan76]

I got error: Stratum authorization when i tried to mine AION on many pool and got 100% reject shares :<. Can you check and fix this. Its like this is for solo mining only T.T

Same problem here.

[invizion]

I got error: Stratum authorization when i tried to mine AION on many pool and got 100% reject shares :<. Can you check and fix this. Its like this is for solo mining only T.T

Same problem here.

You are likely using a wrong address or specifying the wrong pers value.

Only mainnet Aion is mineable, addresses start with 0xa and you will need to download and generate such an address using the Aion Kernel. I see many people trying to mine with an Ethereum Address, the Aion ERC20 token is not mineable.

The personalisation line is optional (recommend leaving it out), if you do include it make sure it's correct: AION0PoW otherwise you'll get errors.

[bullout]

Anyone know of a watcher program that works with 0.3 - 0.4 ? I was using AcreFawn's FarmWatchBot with 0.3.4b, but it hasn't been updated since the new release. I used to have 99.9% uptime but now my rig will hang up and the program doesn't recognize the error so it doesn't restart the miner. https://bitcointalk.org/index.php?topic=2071108.0

[Regulater]

Anyone know of a watcher program that works with 0.3 - 0.4 ? I was using AcreFawn's FarmWatchBot with 0.3.4b, but it hasn't been updated since the new release. I used to have 99.9% uptime but now my rig will hang up and the program doesn't recognize the error so it doesn't restart the miner. https://bitcointalk.org/index.php?topic=2071108.0

Yeah Awesome Miner

[cryptdough5]

Why is it not possible to mine this with AMD cards? Might be a stupid question but I'd like to understand how that works?

[ComputerGenie]

Why is it not possible to mine this with AMD cards? Might be a stupid question but I'd like to understand how that works?

Because it's like trying to put Chevy parts in a Ford; they just aren't made the same and it takes special items to be able to fit both.

[Zheko87]

How many sol/s on 144_5 algorithm can I squeeze with 1060 in overclocking?

I got 203  Sol/s on 9x1060 #GB with micron memory, 22.5 average (140 core, 1400 memory and 80W settings in simplemining.net). Add 10% for 6GB I guess. But ETH still pays better than BTG for me.

ETH may be better then BTG, but some 144_5 crypto gives more Testing 0.4 version on 6x1066 gives me ~150 sol/s with light overlock core +100 only

[hell_slayer]

Is there any way to decrease mining intensity ? Seems like argument --intensity and modification of intensity parameter in cfg file didn't work for me   I would like to decrease intensity by the half to be able to play some games while mining on my working PC  

[ComputerGenie]

Is there any way to decrease mining intensity ?...

Is there any way to get people to read a thread instead of asking the same question that's, literally, been asked 100 times and answered by dev on the exact same page the newest ask is on?

FFS, it's not like you even had to dig through every page (but rather just 12 posts back).