neoscrypt algo for ccminer and nvidia card: call for donation

[IncludeBeer]

Anyone else using Coinotron see a discrepancy with hash rate?  With the latest miner from djm34, I get ~630khz on my gtx 980.  However, Coinotron consistently shows 270-300khz.  Usually you only see this big of a difference with very low hash rates.  I was just curiousif this is a known issue, and/or what the solution is.  Thanks!

[djm34]

Anyone else using Coinotron see a discrepancy with hash rate?  With the latest miner from djm34, I get ~630khz on my gtx 980.  However, Coinotron consistently shows 270-300khz.  Usually you only see this big of a difference with very low hash rates.  I was just curiousif this is a known issue, and/or what the solution is.  Thanks!

may-be you can try to use --diff 0.5  not sure it will really help, it can be that you are requesting share at a certain level of difficulty and miss all the share at lower difficulty.

[IncludeBeer]

Anyone else using Coinotron see a discrepancy with hash rate?  With the latest miner from djm34, I get ~630khz on my gtx 980.  However, Coinotron consistently shows 270-300khz.  Usually you only see this big of a difference with very low hash rates.  I was just curiousif this is a known issue, and/or what the solution is.  Thanks!

may-be you can try to use --diff 0.5  not sure it will really help, it can be that you are requesting share at a certain level of difficulty and miss all the share at lower difficulty.

Sorry, forgot to include my config:
ccminer.exe -a neoscrypt --throughput 3 --diff 0.5 -o stratum+tcp://coinotron.com:3337 -u NewbieMiner.gpu -p derpy

If I don't use the --diff 0.5, i get boo's about 2/3 the time. I figure it might be something to do with difficulty and how Coinotron calculates hashrate on their side.  Perhaps with this setting diff to 0.5, my shares are worth less, therefore they calculate my hashrate to be less?  I'm really just pullin things outta my ass here; it's all conjecture.

[djm34]

Anyone else using Coinotron see a discrepancy with hash rate?  With the latest miner from djm34, I get ~630khz on my gtx 980.  However, Coinotron consistently shows 270-300khz.  Usually you only see this big of a difference with very low hash rates.  I was just curiousif this is a known issue, and/or what the solution is.  Thanks!

may-be you can try to use --diff 0.5  not sure it will really help, it can be that you are requesting share at a certain level of difficulty and miss all the share at lower difficulty.

Sorry, forgot to include my config:
ccminer.exe -a neoscrypt --throughput 3 --diff 0.5 -o stratum+tcp://coinotron.com:3337 -u NewbieMiner.gpu -p derpy

If I don't use the --diff 0.5, i get boo's about 2/3 the time. I figure it might be something to do with difficulty and how Coinotron calculates hashrate on their side.  Perhaps with this setting diff to 0.5, my shares are worth less, therefore they calculate my hashrate to be less?  I'm really just pullin things outta my ass here; it's all conjecture.

yes the --diff 0.5 is linked to the way they set their difficulty. The share aren't worht less, you may just get less shares than you should.
It is a bit tricky to set.
I tried to use coinotron some times ago and I noticed that very large pool reports fantasist hashrate for small miners... also they don't seem to do diff retargetting meaning all your shares have the same value.

so unless you have a huge farm, it would probably better to go to an other pool.
May-be you could try yamp or smaller pool... (blockfactory is the one I use, hashrate is well reported... never saw any big discrepancies in hashrate)

[tflmike]

Hi

Just sent my donation for the Windows version ccMiner NeoScrypt, Nice Work keep it up

Net amount: 0.03000262 BTC
Transaction ID: 3e1819bc7dc5abe9cd5d80bcfcb81a82b7e8d08cf43753e7fe49f77af545480d-000

[IncludeBeer]

Anyone else using Coinotron see a discrepancy with hash rate?  With the latest miner from djm34, I get ~630khz on my gtx 980.  However, Coinotron consistently shows 270-300khz.  Usually you only see this big of a difference with very low hash rates.  I was just curiousif this is a known issue, and/or what the solution is.  Thanks!

may-be you can try to use --diff 0.5  not sure it will really help, it can be that you are requesting share at a certain level of difficulty and miss all the share at lower difficulty.

Sorry, forgot to include my config:
ccminer.exe -a neoscrypt --throughput 3 --diff 0.5 -o stratum+tcp://coinotron.com:3337 -u NewbieMiner.gpu -p derpy

If I don't use the --diff 0.5, i get boo's about 2/3 the time. I figure it might be something to do with difficulty and how Coinotron calculates hashrate on their side.  Perhaps with this setting diff to 0.5, my shares are worth less, therefore they calculate my hashrate to be less?  I'm really just pullin things outta my ass here; it's all conjecture.

yes the --diff 0.5 is linked to the way they set their difficulty. The share aren't worht less, you may just get less shares than you should.
It is a bit tricky to set.
I tried to use coinotron some times ago and I noticed that very large pool reports fantasist hashrate for small miners... also they don't seem to do diff retargetting meaning all your shares have the same value.

so unless you have a huge farm, it would probably better to go to an other pool.
May-be you could try yamp or smaller pool... (blockfactory is the one I use, hashrate is well reported... never saw any big discrepancies in hashrate)

Thanks for the feedback!  I made an account at blocksFactory and let it run ~12 hours.

As can be seen, the miner is still rockin at ~630 kh/s, the difficulty they show for the worker is 64, and the "3 hour", "12 hour" reported hashrates are consistent with what I was seeing at Coinotron.  I'll give it some more time and see if that kh/s reporting gap doesn't get closed a bit.

[djm34]

Anyone else using Coinotron see a discrepancy with hash rate?  With the latest miner from djm34, I get ~630khz on my gtx 980.  However, Coinotron consistently shows 270-300khz.  Usually you only see this big of a difference with very low hash rates.  I was just curiousif this is a known issue, and/or what the solution is.  Thanks!

may-be you can try to use --diff 0.5  not sure it will really help, it can be that you are requesting share at a certain level of difficulty and miss all the share at lower difficulty.

Sorry, forgot to include my config:
ccminer.exe -a neoscrypt --throughput 3 --diff 0.5 -o stratum+tcp://coinotron.com:3337 -u NewbieMiner.gpu -p derpy

If I don't use the --diff 0.5, i get boo's about 2/3 the time. I figure it might be something to do with difficulty and how Coinotron calculates hashrate on their side.  Perhaps with this setting diff to 0.5, my shares are worth less, therefore they calculate my hashrate to be less?  I'm really just pullin things outta my ass here; it's all conjecture.

yes the --diff 0.5 is linked to the way they set their difficulty. The share aren't worht less, you may just get less shares than you should.
It is a bit tricky to set.
I tried to use coinotron some times ago and I noticed that very large pool reports fantasist hashrate for small miners... also they don't seem to do diff retargetting meaning all your shares have the same value.

so unless you have a huge farm, it would probably better to go to an other pool.
May-be you could try yamp or smaller pool... (blockfactory is the one I use, hashrate is well reported... never saw any big discrepancies in hashrate)

Thanks for the feedback!  I made an account at blocksFactory and let it run ~12 hours.

As can be seen, the miner is still rockin at ~630 kh/s, the difficulty they show for the worker is 64, and the "3 hour", "12 hour" reported hashrates are consistent with what I was seeing at Coinotron.  I'll give it some more time and see if that kh/s reporting gap doesn't get closed a bit.

Yeah that's strange after 3hours... May-be try to rise the difficulty; 64 isn't a lot try to increase a bit.
I always had my hashrate reported rather well by blockfactory

[blue-yu]

Any new versions, development? Or should we pay more?

[djm34]

Any new versions, development? Or should we pay more?

Hi, not for the moment, I have been working on some other algo.
Also neoscrypt was released after several optimizations were applied, so it is already pretty much optimized

It is probably possible to increase even more the performance (8 way algo) but would require a complete rewriting on most of the code...) may-be later

but if enough fund can be found, I can look again

[chrysophylax]

Any new versions, development? Or should we pay more?

Hi, not for the moment, I have been working on some other algo.
Also neoscrypt was released after several optimizations were applied, so it is already pretty much optimized

It is probably possible to increase even more the performance (8 way algo) but would require a complete rewriting on most of the code...) may-be later

djm - would you release the source ( even privately if not publicly ) so that i can compile natively please? ...

that would be very much appreciated - and as always - will be very appreciative ...

it seems i am always stuck with the compiled miners after i have donated only to have them not work on my systems anyway ...

i have no issue mining as an extra donation to your address ( using the miner of course ) as a tanx - as i always do ...

pm me if you have any pre-requisites to this ...

#crysx

[djm34]

Any new versions, development? Or should we pay more?

Hi, not for the moment, I have been working on some other algo.
Also neoscrypt was released after several optimizations were applied, so it is already pretty much optimized

It is probably possible to increase even more the performance (8 way algo) but would require a complete rewriting on most of the code...) may-be later

djm - would you release the source ( even privately if not publicly ) so that i can compile natively please? ...

that would be very much appreciated - and as always - will be very appreciative ...

it seems i am always stuck with the compiled miners after i have donated only to have them not work on my systems anyway ...

i have no issue mining as an extra donation to your address ( using the miner of course ) as a tanx - as i always do ...

pm me if you have any pre-requisites to this ...

#crysx

I don't plan to release the source at the moment and as long as there will be people asking for it so that they don't donate for nothing (it wouldn't be fair to them)

[thefix]

I just sent some Bitcoin to help with the fund and PMed you the transaction ID, I would love to get your miner when you get a chance.

[blue-yu]

Just switched from Linux to Windows mining (mainly because of Miner Control under Windows). Mining Neoscrypt under Linux was rock solid, but under Windows is not so good. On my 6*750ti rig  I am using "--throughput 3,3,3,3,3,3 --diff 0.5" config, but had a lot of rejects with Yaamp. Another problem is that in the middle of mining one or few GPU's crush, so reboot is only solution.

So, do we need a new version of ccminer?

BR...

[djm34]

Just switched from Linux to Windows mining (mainly because of Miner Control under Windows). Mining Neoscrypt under Linux was rock solid, but under Windows is not so good. On my 6*750ti rig  I am using "--throughput 3,3,3,3,3,3 --diff 0.5" config, but had a lot of rejects with Yaamp. Another problem is that in the middle of mining one or few GPU's crush, so reboot is only solution.

So, do we need a new version of ccminer?

BR...

I think you should try to identify which card misbehave, then try to either decrease the OC or the throughput for these cards.

Regarding rejected shares, try to vary the diff, just in case (don't remember if it is exactly the same) say  2, 0.5 and 0.25.
also make sure you have a pagefile large enough (with 6 cards, the pagefile must be rather large, it is possible to check the graphics for all that with msi AB)

[thefix]

Just switched from Linux to Windows mining (mainly because of Miner Control under Windows). Mining Neoscrypt under Linux was rock solid, but under Windows is not so good. On my 6*750ti rig  I am using "--throughput 3,3,3,3,3,3 --diff 0.5" config, but had a lot of rejects with Yaamp. Another problem is that in the middle of mining one or few GPU's crush, so reboot is only solution.

So, do we need a new version of ccminer?

BR...

Try --throughput 1

I run 6 x750ti and that works perfectly on all cards, anything above that and things crash

[blue-yu]

Try --throughput 1

I run 6 x750ti and that works perfectly on all cards, anything above that and things crash

Do You use it with diff parameter or not?

[surfrider]

just sent 0.03 donation.
transaction id: 31dc6852876d04836bb563d964454eb995d05ab10d398d6a6f77c69b6301be39

windows ccminer please.

[Slava_K]

Sent 0.04 BTC.
Windows please...
transactio ID 658dba77e2bf084cf115fadd15fd6be6b93438711f5d8bffcce9155453907187

[thefix]

Try --throughput 1

I run 6 x750ti and that works perfectly on all cards, anything above that and things crash

Do You use it with diff parameter or not?

I do not use diff parameter and I have been able to solo mine, pool mine and run on nicehash very stable.

Occasionally it will crash, but not often

[scryptr]

DJM34 --

This thread has slowed some, but I hope readers still contribute.  I just sent another two-bits worth your way today.  I am hoping that the funding completes, maybe sp_ could crank it up even more. 

I discovered that with "--throughput 1", and no other additional flags, my GTX SSC 960 will hash at a reliable 325kh/s on Win 7 x64.  That is slightly better than a 780ti.  I have a couple GTX 970 FTW on the way, and will holler back when I get some data to report.       --scryptr