[ANN] DelosMiner for nVidia GPUs by TeamDelos

[sonkub]

Can't benchmark with hsr,x16r Algorithm.

32-bit or 64-bit? Which CUDA version? Can you post an error msg.?

Can't benchmark with hsr,x16r Algorithm.

Please try the latest release 1.2.2, which is adressing stability issues for x16r and other algos!

ccminer show blakecoin only.

ccminer version Win 64-bit with CUDA 9.1

I try version 1.2.2 but same.

command

Code:

ccminer --benchmark --algo=x16r

https://preview.ibb.co/gsnciJ/benchmark.png

[somaton]

Thanks for 9.1 builds. Tried Phi with 1.2.2 32 and 64-bit - from time to time got message "result does not validate on CPU" and after 5 minutes driver 390.77 crashed. Hsr still needs some work, crashing easily and slow compared to other miners.

[TeamDelos]

Thanks for 9.1 builds. Tried Phi with 1.2.2 32 and 64-bit - from time to time got message "result does not validate on CPU" and after 5 minutes driver 390.77 crashed. Hsr still needs some work, crashing easily and slow compared to other miners.

Thanks for reporting this! Haven't seen this behavior recently. But we'll definetly have a look into it!!!

[Quake4]

1.2.2 9.2 32bit
Api isnt answer on request "threads". In 9.1 all fine.

Please, move fee time to 5 min and above from miner start. Wrong speed detect in autoswitch software.

[josywong]

1.2.2 9.2 32bit
Api isnt answer on request "threads". In 9.1 all fine.

Please, move fee time to 5 min and above from miner start. Wrong speed detect in autoswitch software.

+1

1.2.2 9.2 32bit crash too much.

[buxstore]

  you  should improve your software like enemy              eg:  - Revised dev-fee mechanism. Now the miner does not break off connection with a pool and does not drop work of video cards. Everything works without switch-offs.

[ivakar]

how about x17 algo? enemy miner has it? this algo is very profitable now so it would be nice if you gonna implement the support of this algo to your miner
by the way II have test your miner for X16r, raven.. still enemy has the lead)) so you have to do more optimization..

[sandinthebones]

Any plan for linux verison?

[PostJok]

Good Hashrates on Phi, but you have to do something with the DevFee. If i optimize my Rig for Phi and it switches to x16r for your Fee, the Driver crashes as it isnt configured for the other Algorithm. Maybe make it Algo depended.

[mo35]

Request , could you please move dev fee away from intial miner launch 10+ mins , since you hardcoded x16r it messes up bechmarkings for other algos. Or even better as people already requested , make it per algo or somethings. current implementation is major annoyance.

Also there is something weird with hashrate reporting after dev fee , on pool it shows correct values , but miner output is wrong , messes up with autoswitching software when its update hashrate statistics periodically.

[TeamDelos]

Request , could you please move dev fee away from intial miner launch 10+ mins , since you hardcoded x16r it messes up bechmarkings for other algos. Or even better as people already requested , make it per algo or somethings. current implementation is major annoyance.

Also there is something weird with hashrate reporting after dev fee , on pool it shows correct values , but miner output is wrong , messes up with autoswitching software when its update hashrate statistics periodically.

https://preview.ibb.co/ei7cq8/delos.png

Thanks for suggesting this! Pushing dev time to much back is an issue as many switching tools (MindMiner, All-In-Miner, etc.) tend to switch quite often (every 2-3 minutes).

Hence, we take your idea and avoid the x16r hardcoding. This should avoid messing up benchmarks of other algos!

Stay tuned for the next release...

[xlth]

How about Xevan?

[TeamDelos]

how about x17 algo? enemy miner has it? this algo is very profitable now so it would be nice if you gonna implement the support of this algo to your miner
by the way II have test your miner for X16r, raven.. still enemy has the lead)) so you have to do more optimization..

Thanks for your feedback!

x17 is already implenented. Does it not work for you?
How much performance are you missing on x16r with which GPU?
We are currently implementing Xevan. Hopefully that is interesting for you, too!

[abdullahsurati]

Can you bring support for P102-100 mining edition? It uses CUDA 6.1

[TeamDelos]

Can you bring support for P102-100 mining edition? It uses CUDA 6.1

This will be probably not possible for all supoorted algo. Which algos are you going to use?

[TeamDelos]

DelosMiner v1.3.0 is released!

Change log:

    Performance boosts on LYRA2v2, Skein and x16r!
    Additional core speed improvements of up to 2% for most algos.
    Revised dev fee mechanism for a seemless mining experience.

Enjoy!

[HashAuger]

DelosMiner v1.3.0 is released!

Change log:

    Performance boosts on LYRA2v2, Skein and x16r!
    Additional core speed improvements of up to 2% for most algos.
    Revised dev fee mechanism for a seemless mining experience.

Thanks for the update. The optimizations to Lyra2v2 are much appreciated.  I have small suggest concerning the dev fee mechanism, however. Would you consider hiding the accepted results and hash rate messages when the miner is working on the dev pool?  The issue is that if you are trying to do any calculations based on those numbers, the results from the dev pool work can skew the averages. Since the dev pool work doesn't correspond to the user's work, it would be preferable to hide performance messages related to the dev pool. Thanks for considering my suggestion.

[grendel25]

How about Xevan?

+1 to this.  Any plans to add Xevan into the optimization mix here?

[mo35]

DelosMiner v1.3.0 is released!

Change log:

    Performance boosts on LYRA2v2, Skein and x16r!
    Additional core speed improvements of up to 2% for most algos.
    Revised dev fee mechanism for a seemless mining experience.

Thanks for the update. The optimizations to Lyra2v2 are much appreciated.  I have small suggest concerning the dev fee mechanism, however. Would you consider hiding the accepted results and hash rate messages when the miner is working on the dev pool?  The issue is that if you are trying to do any calculations based on those numbers, the results from the dev pool work can skew the averages. Since the dev pool work doesn't correspond to the user's work, it would be preferable to hide performance messages related to the dev pool. Thanks for considering my suggestion.

i strongly support this , just simple "starting devfee mining" and "finishing mining devfee" messages should be sufficient for thos who want to see time stamps or whatsnot

[fubarmining]

DelosMiner v1.3.0 is released!

Change log:

    Performance boosts on LYRA2v2, Skein and x16r!
    Additional core speed improvements of up to 2% for most algos.
    Revised dev fee mechanism for a seemless mining experience.

Thanks for the update. The optimizations to Lyra2v2 are much appreciated.  I have small suggest concerning the dev fee mechanism, however. Would you consider hiding the accepted results and hash rate messages when the miner is working on the dev pool?  The issue is that if you are trying to do any calculations based on those numbers, the results from the dev pool work can skew the averages. Since the dev pool work doesn't correspond to the user's work, it would be preferable to hide performance messages related to the dev pool. Thanks for considering my suggestion.

i strongly support this , just simple "starting devfee mining" and "finishing mining devfee" messages should be sufficient for thos who want to see time stamps or whatsnot

I strongly support this also.  Can you also remove the dev pool work from the api.  It is throwing off the miners calculated stats.