CryptoDredge 0.27.0 — CUDA Miner

[cryptoyes]

... it's getting worse, it feels as if some cache of dupe shares is growing with every devfee session or something. heres the latest devfee session and all accepted/rejected shares after it (1503/23 is now 1.5% wasted hashrate)

Code:

[23:43:26] INFO  - Start devfee session
[23:44:00] INFO  - Start user session
[23:44:17] INFO  - 1477/14 Accepted : diff=2514565 : 70.15KH/s
[23:44:29] INFO  - 1478/14 Accepted : diff=4254965 : 70.15KH/s
[23:44:31] INFO  - 1478/15 Rejected : diff=3107198, reason: Duplicate share2
[23:45:24] INFO  - 1479/15 Accepted : diff=9943071 : 70.15KH/s
[23:45:33] INFO  - 1480/15 Accepted : diff=1526505 : 70.15KH/s
[23:45:34] INFO  - 1480/16 Rejected : diff=1995470, reason: Duplicate share2
[23:45:49] INFO  - 1481/16 Accepted : diff=6100783 : 70.15KH/s
[23:45:50] INFO  - 1482/16 Accepted : diff=3564395 : 70.15KH/s
[23:46:02] INFO  - 1483/16 Accepted : diff=9404347 : 70.15KH/s
[23:46:47] INFO  - 1484/16 Accepted : diff=1574540 : 70.15KH/s
[23:47:10] INFO  - 1485/16 Accepted : diff=4827008 : 70.15KH/s
[23:47:48] INFO  - 1485/17 Rejected : diff=1707333, reason: Duplicate share2
[23:47:51] INFO  - 1485/18 Rejected : diff=45380214, reason: Duplicate share2
[23:48:10] INFO  - 1486/18 Accepted : diff=5513997 : 70.15KH/s
[23:48:32] INFO  - 1487/18 Accepted : diff=3738519 : 70.15KH/s
[23:48:41] INFO  - 1488/18 Accepted : diff=16101789 : 70.15KH/s
[23:49:35] INFO  - 1489/18 Accepted : diff=2454525 : 70.15KH/s
[23:49:50] INFO  - 1490/18 Accepted : diff=2369070 : 70.15KH/s
[23:50:08] INFO  - 1491/18 Accepted : diff=5169119 : 70.15KH/s
[23:50:19] INFO  - 1491/19 Rejected : diff=31286175, reason: Duplicate share2
[23:50:35] INFO  - 1492/19 Accepted : diff=7178882 : 70.15KH/s
[23:50:37] INFO  - 1492/20 Rejected : diff=1798143, reason: Duplicate share2
[23:51:19] INFO  - 1492/21 Rejected : diff=3132001, reason: Duplicate share2
[23:51:50] INFO  - 1493/21 Accepted : diff=9671138 : 70.15KH/s
[23:51:53] INFO  - 1494/21 Accepted : diff=2608553 : 70.15KH/s
[23:52:14] INFO  - 1495/21 Accepted : diff=1965863 : 70.15KH/s
[23:52:17] INFO  - 1496/21 Accepted : diff=2277337 : 70.15KH/s
[23:52:29] INFO  - 1496/22 Rejected : diff=1741662, reason: Duplicate share2
[23:52:47] INFO  - 1497/22 Accepted : diff=1929785 : 70.15KH/s
[23:53:19] INFO  - 1498/22 Accepted : diff=5289919 : 70.15KH/s
[23:53:20] INFO  - 1499/22 Accepted : diff=2230582 : 70.15KH/s
[23:53:26] INFO  - 1500/22 Accepted : diff=2419414 : 70.15KH/s
[23:53:50] INFO  - 1501/22 Accepted : diff=1556715 : 70.15KH/s
[23:54:08] INFO  - 1501/23 Rejected : diff=2105250, reason: Duplicate share2
[23:54:13] INFO  - 1502/23 Accepted : diff=1619847 : 70.15KH/s
[23:54:31] INFO  - 1503/23 Accepted : diff=22358617 : 70.15KH/s

[cryptoyes]

Here's the entire log of sessions and shares: https://pastebin.com/s97g6ps9

You can see which shares are duplicated by the diff value. For example, share 19 is a duplicate of share 11, and share 229 is a duplicate of share 224, and so on

I can't work out what exactly causes them, but they always pop-up after the devfee and they are all a duplicate of some older sahre (not the immediately previous one).

Could you investigate and fix this?

[cryptoyes]

and here's another "duplicate share" (the previous ones were "duplicate share2" ... note the "2", that's probably from the pool?) which got me banned: https://pastebin.com/cp3dPSiu

... again, happened shortly after a devfee session. Please look into this, it sounds like it should be an easy fix

[tbearhere]

       Also have this error on ver 23 mtp-tcr.

[mstrozier]

I've had a chance to run this over my rig and it'll work on all old and modern Nvidia cards except for the 1650 Super.  It only gives a GPU Critical Error message and restarts the miner.  Works fine on all other cards ranging from a 750 ti, 1050ti, 1660ti, 1070 and RTX 2080.  Just the 1650 Super fails. 

Since this GPU is still relatively new, has it not been coded for yet?  I am currently running the latest version 23.0

[JayDDee]

       Also have this error on ver 23 mtp-tcr.

Stale shares can't be complety avoided, there will alway be a window where a submitted share or 2 can overlap with a
new job. High network latency and low hashrate algos (longer time/thread to calculate 1 hash) make stales more likely.
In this case it looks like it may be related to the fee switch (something I have no experience with).

[JayDDee]

I've had a chance to run this over my rig and it'll work on all old and modern Nvidia cards except for the 1650 Super.  It only gives a GPU Critical Error message and restarts the miner.  Works fine on all other cards ranging from a 750 ti, 1050ti, 1660ti, 1070 and RTX 2080.  Just the 1650 Super fails. 

Since this GPU is still relatively new, has it not been coded for yet?  I am currently running the latest version 23.0

Update driver?

[tbearhere]

       Also have this error on ver 23 mtp-tcr.

Stale shares can't be complety avoided, there will alway be a window where a submitted share or 2 can overlap with a
new job. High network latency and low hashrate algos (longer time/thread to calculate 1 hash) make stales more likely.
In this case it looks like it may be related to the fee switch (something I have no experience with).

It seems it is related to the fee switch. Never happens with older versions.

Also if we update drivers the cpu will run at 100% load.

[mstrozier]

I've had a chance to run this over my rig and it'll work on all old and modern Nvidia cards except for the 1650 Super.  It only gives a GPU Critical Error message and restarts the miner.  Works fine on all other cards ranging from a 750 ti, 1050ti, 1660ti, 1070 and RTX 2080.  Just the 1650 Super fails. 

Since this GPU is still relatively new, has it not been coded for yet?  I am currently running the latest version 23.0

Update driver?

I was at 441 on the nvidia driver.  Updated to 442.74  and same issue.  Does not hash and states critical issue and restarts the miner.   I can use this card on any other mining software, claymore 15 on coins that'll allow due to dag size (callisto, Ether-1 as examples) without any problem.  I can use T-Rex on Ravencoin again no problem.  Card works great.  However if I use Cryptodredge it'll bomb out every time.  So now with everything updated just leads me to thinking something in the mining software.  Since the 1650 Super is fairly new and not many people using the card.

[JayDDee]

I was at 441 on the nvidia driver.  Updated to 442.74  and same issue.  Does not hash and states critical issue and restarts the miner.   I can use this card on any other mining software, claymore 15 on coins that'll allow due to dag size (callisto, Ether-1 as examples) without any problem.  I can use T-Rex on Ravencoin again no problem.  Card works great.  However if I use Cryptodredge it'll bomb out every time.  So now with everything updated just leads me to thinking something in the mining software.  Since the 1650 Super is fairly new and not many people using the card.

Could be a tuning issue. Does Cryptodredge have an intensity option like ccminer? Try lowering it until the SW is updated to
properly tune for that card.

[mstrozier]

I was at 441 on the nvidia driver.  Updated to 442.74  and same issue.  Does not hash and states critical issue and restarts the miner.   I can use this card on any other mining software, claymore 15 on coins that'll allow due to dag size (callisto, Ether-1 as examples) without any problem.  I can use T-Rex on Ravencoin again no problem.  Card works great.  However if I use Cryptodredge it'll bomb out every time.  So now with everything updated just leads me to thinking something in the mining software.  Since the 1650 Super is fairly new and not many people using the card.

Could be a tuning issue. Does Cryptodredge have an intensity option like ccminer? Try lowering it until the SW is updated to
properly tune for that card.

Thanks, didn't think about the tuning.  That did the trick.  It fails on 6 however 5 and below it works great.   Appreciate it.

[Eketubol]

Is there any reason why MTP is still at 2% fee? I would probably use CryptoDredge if the miner outperforms its competitors but at the moment T-rex Miner works better on MTP for less fee. Can you please update the fee to reflect the current mining climate.

[ZeeeN]

Is there any reason why MTP is still at 2% fee? I would probably use CryptoDredge if the miner outperforms its competitors but at the moment T-rex Miner works better on MTP for less fee. Can you please update the fee to reflect the current mining climate.

for MTP you can use ccminer with same speed

[igotek]

2020-07-20Announce Haven hardfork succeeded successfully.
We now have two different payment systems. XHV and xUSD.
You can set minimum payment amounts here.
Please do not forget to update miners and wallet.
2020-07-15AnnounceHaven launching xUSD on July 20th.
Miner algo stays the same cn-heavy/xhv but need to update will be required.
Please do not forget to update miners and wallet.
You can find all working miners here.


Any update for XHV ?

[cryptoyes]

Haven already forked. Any updated version to support it?

[coinminery]

https://prnt.sc/tqfgvq
https://prnt.sc/tqfgvq

[cryptodredge]

v0.24.0 released

• Improve Nimiq algorithm
• Add -q, --recompute option to set memory reduction factor (1 - 32) (Nimiq only). For example: -q N[,N] (default: 23)
• Fix high CPU usage on MTP-like, Argon2-like and CryptoNight-like algorithms
• Slightly improve Chukwa (Argon2-512) and Ninja (Argon2-256) algorithms
• MTP: Fix memory leaks
• New CryptoNightTLO algorithm
• Fix CryptoNightHaven issue related to 'Unsupported blob size'
• New Sha256csm algorithm
• Add --sni option to enable SNI (Server Name Indication) for the pool connection
• Other minor fixes

Download: https://github.com/technobyl/CryptoDredge/releases/tag/v0.24.0

Example:

Code:

CryptoDredge -a argon2d-nim -o wss://nimiq.icemining.ca:2053 -u "NQ68 XM08 9BE8 57DE QMT6 PAT5 PJGQ TEBX FSBU" -p "My Miner" -q 21

Happy mining!

[tbearhere]

Intensity settings aren't working or should I say - i 0 still isn't low enough to mine and use my computer too. This also happens in v0.23.0 release too. Windows 8.1

[mirny]

v0.24.0 released

• Improve Nimiq algorithm
• Add -q, --recompute option to set memory reduction factor (1 - 32) (Nimiq only). For example: -q N[,N] (default: 23)
• Fix high CPU usage on MTP-like, Argon2-like and CryptoNight-like algorithms
• Slightly improve Chukwa (Argon2-512) and Ninja (Argon2-256) algorithms
• MTP: Fix memory leaks
• New CryptoNightTLO algorithm
• Fix CryptoNightHaven issue related to 'Unsupported blob size'
• New Sha256csm algorithm
• Add --sni option to enable SNI (Server Name Indication) for the pool connection
• Other minor fixes

Download: https://github.com/technobyl/CryptoDredge/releases/tag/v0.24.0

Example:

Code:

CryptoDredge -a argon2d-nim -o wss://nimiq.icemining.ca:2053 -u "NQ68 XM08 9BE8 57DE QMT6 PAT5 PJGQ TEBX FSBU" -p "My Miner" -q 21

Happy mining!

Hi, could you please compile binary for cuda 9.1(linux)?
cheers

[arielbit]

MTP latest nvidia drivers.

windows 7 ~3MH
windows10 ~4MH