WildRig: new miner for Wild Keccak algo

[badfad]

5.7Mh/s @ 140 W Vega56, HBM temp @ 74 C.

Can it do more? I remember getting 7.5Mh/s when purk launched but scratchpad has grown since then.

Also pool reports a stable 5.4-6.0Mh/s, unlike the official miner 

[1714735922]

1714735922

[locbet]

How to set static diff on miner tool

[andrucrypt]

How to set static diff on miner tool

this depends on pool you are using. Some pools can accept it via appending static diff to wallet address like this: --user YOUR_WALLET.STATIC_DIFF

[krjalone]

Are we close yet to launch new miner version? 

[igotek]

One of the GPU doesnt work. Different GPU doesnt work on each new start of the miner. WHY ?
All of the GPUs work on SRB Miner without any problem.

And how could I see the GPU temp ?

AMD RX570 4GB GPUs are moded for cryptonight-v7.
is it possible to get higher hashrate ?

Picture 1:



Picture 2:


I have 402 GPU. if boolerry is more profitable and stable. i will switch my farm to your miner.

Code:

@echo off

:loop
wildrig.exe --opencl-launch=256x0 --send-stale --url=bbr.luckypool.io:5577 --user=bbRAGq8ZjBqQFUNdyWzwixE7RPB9LskWuH1viYvQShpR6wqMZfZ35LnEZ5tVizHAB7cNK4r6yBynYUAJCjDuHtLZJn4skkZ3gwWZMYBGtUwQxcJumHgX1F38vL16NivE3XBcAdJc9oGP3D --pass=YA6 --scratchpad-file=scratchpad.bin --scratchpad-url=http://eu-bbr.luckypool.io/scratchpad.bin
if ERRORLEVEL 1000 goto custom
timeout /t 5
goto loop

:custom
echo Miner exited because of some error
timeout /t 5
goto loop

[andrucrypt]

One of the GPU doesnt work. Different GPU doesnt work on each new start of the miner. WHY ?

This algo is pretty hard and wants more stability from GPU. Will contact you DM.

Are we close yet to launch new miner version? 

Probably tomorrow I will release it

[igotek]

@andrucrypt

Thank u for your quick help.

**

Reference for other new users:

Rigs have moded bios for cryptonight-v7 (overclocked and undervolted)

12 x AMD RX470 4GB Saphire Mining Edition 11650 h/s for cryptonight-v7 (SRBMiner v1.6.6 - intensity 59) - 1480W
12 x AMD RX470 4GB Saphire Mining Edition 24300 kh/s for wildkeccak Boolbery (Wildrig-0.10.4-beta default intensity) - 1430W

12 x AMD RX570 4GB - Asus Strix Gaming has 11310 h/s for cryptonight-v7 (SRBMiner v1.6.6 - intensity 59)
12 x AMD RX570 4GB - Asus Strix Gaming has 24060 kh/s for wildkeccak Boolbery (Wildrig-0.10.4-beta default intensity)

7.30$ for cryptonight-v7
7.50$ for wildkeccak Boolbery

**

I keep my 2 rigs running wildrig. I can switch my other rigs after checking everything is ok.




I think RX470 is better than RX570.

[Mind163]

The higher the core and memory frequencies are on AMD, the greater the hashed.
Пpи 1250 ~ 2150-2200 Kh/s

[igotek]

...
Download for HiveOS(custom miner integration)
(dependencies: libcurl, libuv1, libmicrohttpd10)
note: use wildrig as a name for miner while installation
...

is it possible to tell me how to install wildrig for HiveOS step by step ?
Only NVIDIA for HiveOS.

[andrucrypt]

Released 0.10.4 beta:
- added new optimized kernels which can improve hashrate for 4-10% on AMD cards
- fixed AMD videocard numeration(was broken in 0.10.3)
- added possibility to set different kernel for each videocard(use parameter --opencl-kernels)
- added parameter --safe-scratchpad-update for rigs with huge amount of gpu's or just not stable motherboards
- [HiveOS]reverted CUDA to 9.0 for compatibility with older drivers

ps: take a look on some notes for Vega and RX570/RX580 cards if you mine BBR.

[tale01]

Released 0.10.4 beta:
- added new optimized kernels which can improve hashrate for 4-10% on AMD cards
- fixed AMD videocard numeration(was broken in 0.10.3)
- added possibility to set different kernel for each videocard(use parameter --opencl-kernels)
- added parameter --safe-scratchpad-update for rigs with huge amount of gpu's or just not stable motherboards
- [HiveOS]reverted CUDA to 9.0 for compatibility with older drivers

ps: take a look on some notes for Vega and RX570/RX580 cards if you mine BBR.

I mine with this version over 6 hour.
Hashrate improve 7% (with kernal 3 and 4). Good job
In terminal console show get job and accept.
Hashrate on pool show is correct.
BUT I not get reward! I mine PURK on purk.fairpool.xyz.
6 Hour for get nothing  

I back to use version 0.9x. It very stable.

Can you test Mine PURK with kern 3 or 4
Thanks

[andrucrypt]

I mine with this version over 6 hour.
Hashrate improve 7% (with kernal 3 and 4). Good job
In terminal console show get job and accept.
Hashrate on pool show is correct.
BUT I not get reward! I mine PURK on purk.fairpool.xyz.
6 Hour for get nothing  

I back to use version 0.9x. It very stable.

Can you test Mine PURK with kern 3 or 4
Thanks

Damn, seems I broke submit method in the last minute... Working on check if everything is correct, but why the heck fairpool doesn't answer back about rejects/etc.

IMPORTANT: EVERYONE, please downgrade until situation is clear.

[andrucrypt]

Reverted back submit method, new builds already available and should work now.

IMPORTANT: EVERYONE, please update to new version 0.10.5

[andrucrypt]

Checked some other pools for PURK and BBR - everything is fine. Anyway, better to update.

Here is some technical details what happened. 0.10.4 version of wildrig was sending nonce as 32-bit value, as string in json it was "728fe500" instead of "728fe50000000000"(64-bit). Hash string was correct all the time. But probably not all pools parse such nonce strings correctly. Currently 0.10.5 sending it again as 64-bit, as was all the time before.

UPDATE: fairpool now should work with 0.10.4 build too, thanks to admin effort. Anyway, still better to update version and check everything.

[tale01]

what does it mean "--send-stale" 

[andrucrypt]

what does it mean "--send-stale" 

If GPU found share right after new job arrived - this share will be "stale". But you can ask miner to send it anyway using this parameter, so pool will decide accept it or not. In most cases it can be accepted.

[tale01]

what does it mean "--send-stale" 

If GPU found share right after new job arrived - this share will be "stale". But you can ask miner to send it anyway using this parameter, so pool will decide accept it or not. In most cases it can be accepted.

Thanks,

Now, I mine with ver 0.10.5
Will report result again 

[andrucrypt]

Thanks,

Now, I mine with ver 0.10.5
Will report result again 

Also, specify please with which build, Windows, Linux or HiveOS?

[tale01]

Thanks,

Now, I mine with ver 0.10.5
Will report result again 

Also, specify please with which build, Windows, Linux or HiveOS?

Windows and HiveOS (not show hashrate)

[igotek]

15 of 18 AMD rigs are updated. (Mining cryptonight with 2 of them. One of them has a problem with wildrig.)
2 of 18 NVIDIA rigs are updated. (I havent installed wildrig to 16 NVIDIA HiveOS rig yet.)

WildRig miner is good. I have been mining for 3 days.

Current issues:
1- HiveOS cannot show the hashrates. I think there is api problem.
2- Asus RX550 2GB has problem with wildrig miner. The miner freeze after several minutes. I couldnt find the solution yet.
3- Asus RX570 4GB Strix Gaming has problem. avg. effort is too high. The other vendors are under %95 - No problem. (Sapphire, MSI, Gigabyte.)

Code:

[15:39:57] accepted(58/14) diff 250.00M avg. effort 130.35% (112 ms)
[15:40:04] accepted(59/14) diff 250.00M avg. effort 129.19% (115 ms)
[15:40:04] accepted(60/14) diff 250.00M avg. effort 127.14% (114 ms)
[15:40:14] accepted(61/14) diff 250.00M avg. effort 126.52% (114 ms)
[15:40:18] accepted(62/14) diff 250.00M avg. effort 125.21% (109 ms)
[15:40:31] accepted(63/14) diff 250.00M avg. effort 125.15% (116 ms)
[15:40:34] speed 10s/60s/15m 24015.9 24020.0 n/a kH/s max: 24074.5 kH/s
[15:40:39] accepted(64/14) diff 250.00M avg. effort 124.36% (117 ms)
[15:40:41] accepted(65/14) diff 250.00M avg. effort 122.77% (116 ms)

165 AMD GPUS and 24 NVIDIA GPUs works well.