CCminer(SP-MOD) Modded GPU kernels.

[joblo]

I've been having a bug that, for whatever reason, happens only in Linux and only in the X11 algo. With that combination, shares are never validated on the CPU. However, Lyra2RE does work, and both algos work flawlessly on Windows. For the record, I'm using a GeForce 750Ti + Intel Core i3 2100 on Arch Linux and Windows 10.

Do you mean you don't get any "yay!!!"s? I've seen this too, only on x11 but it seems the shares
are counted in the pool.

[chrysophylax]

I've been having a bug that, for whatever reason, happens only in Linux and only in the X11 algo. With that combination, shares are never validated on the CPU. However, Lyra2RE does work, and both algos work flawlessly on Windows. For the record, I'm using a GeForce 750Ti + Intel Core i3 2100 on Arch Linux and Windows 10.

cuda version? ...

i have the same issue ( listed a page back and directed at sp ) and compile using cuda 7.0-28 ...

cuda 6.5 has none of these issues - and compiles the miner at a higher hashrate ...

#crysx

[chrysophylax]

I've been having a bug that, for whatever reason, happens only in Linux and only in the X11 algo. With that combination, shares are never validated on the CPU. However, Lyra2RE does work, and both algos work flawlessly on Windows. For the record, I'm using a GeForce 750Ti + Intel Core i3 2100 on Arch Linux and Windows 10.

Do you mean you don't get any "yay!!!"s? I've seen this too, only on x11 but it seems the shares
are counted in the pool.

i get yays - but cpu validation errors as well ... when its compiled with cuda 7.0 ...

this is with x11 and quark ...

#crysx

[sp_]

We are here and working on updates!

You didn't take a day off did you? Naughty person!

We have been working and testing ways to hardfork and make the coin more GPU resistant, along with some major changes to POS 2.0 and it's security flaws.
More news is coming in daytime USA time!

It might not be worth the effort to finish the gpu miner for axiom. The developer is changing the POW algorithm. (from the Axiom thread)

[pallas]

We are here and working on updates!

You didn't take a day off did you? Naughty person!

We have been working and testing ways to hardfork and make the coin more GPU resistant, along with some major changes to POS 2.0 and it's security flaws.
More news is coming in daytime USA time!

It might not be worth the effort to finish the gpu miner for axiom. The developer is changing the POW algorithm. (from the Axiom thread)

I don't know if djm34 is already working on lyra2re2 for ccminer, but if he isn't, that might be an interesting algo to add, instead of axiom.
About neoscrypt, I could get another couple percent of hashrate and I'd like to push it when done.
There was a good amount of code to enhance, remove and cleanup.
But I don't have other cards than 970 so it might be slower or not work at all on them.

[djm34]

We are here and working on updates!

You didn't take a day off did you? Naughty person!

We have been working and testing ways to hardfork and make the coin more GPU resistant, along with some major changes to POS 2.0 and it's security flaws.
More news is coming in daytime USA time!

It might not be worth the effort to finish the gpu miner for axiom. The developer is changing the POW algorithm. (from the Axiom thread)

I don't know if djm34 is already working on lyra2re2 for ccminer, but if he isn't, that might be an interesting algo to add, instead of axiom.

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

[pallas]

I don't know if djm34 is already working on lyra2re2 for ccminer, but if he isn't, that might be an interesting algo to add, instead of axiom.

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

ops sorry, I didn't see it coming, thanks!
will try that later, along with sgminer, provided my wallet does finally sync to testnet, or the testnet pool works.
then maybe sp can merge lyra2re2 into his fork.

[sp_]

I don't know if djm34 is already working on lyra2re2 for ccminer, but if he isn't, that might be an interesting algo to add, instead of axiom.
About neoscrypt, I could get another couple percent of hashrate and I'd like to push it when done.
There was a good amount of code to enhance, remove and cleanup.
But I don't have other cards than 970 so it might be slower or not work at all on them.

I can help you tune your changes on the 750ti. If it's slow we can split it like in djm34's lyra2re code. A seperate kernal for the compute 5.0. Just use github and add a pullrequest.

[djm34]

I don't know if djm34 is already working on lyra2re2 for ccminer, but if he isn't, that might be an interesting algo to add, instead of axiom.

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

ops sorry, I didn't see it coming, thanks!
will try that later, along with sgminer, provided my wallet does finally sync to testnet, or the testnet pool works.
then maybe sp can merge lyra2re2 into his fork.

if you want, I can start my wallet, I have no problem syncing

[sp_]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

[djm34]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

well it is still in their best interest... since exchanges, pool will switch
(and it won't be the case anymore at time of the fork and they might end-up on their own fork, considering wallet will have a new check point set)

actually they will probably mine the last block and be out without noticing

ps: the situation was much worst when they forked out the asic miners

[zjy]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

any coin Profit now 

[lawrencelyl]

Looks like quark coin continue to do down 

[sp_]

Looks like quark coin continue to do down 

0.7BTC/GHASH

http://yiimp.ccminer.org/

[Epsylon3]

too late... no more buy orders and "only" 2400 BTCRY per block

but was a good find like anime ^^ was more 0.5BTC / 10MH or even MH since we got all the blocks during 24h

yes... sometimes a Satcoin can do incomes... when you dont share the info

[bensam1231]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

well it is still in their best interest... since exchanges, pool will switch
(and it won't be the case anymore at time of the fork and they might end-up on their own fork, considering wallet will have a new check point set)

actually they will probably mine the last block and be out without noticing

ps: the situation was much worst when they forked out the asic miners

Scrypt-N had ASICs?

[djm34]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

well it is still in their best interest... since exchanges, pool will switch
(and it won't be the case anymore at time of the fork and they might end-up on their own fork, considering wallet will have a new check point set)

actually they will probably mine the last block and be out without noticing

ps: the situation was much worst when they forked out the asic miners

Scrypt-N had ASICs?

sure... still has

[flipclip]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

well it is still in their best interest... since exchanges, pool will switch
(and it won't be the case anymore at time of the fork and they might end-up on their own fork, considering wallet will have a new check point set)

actually they will probably mine the last block and be out without noticing

ps: the situation was much worst when they forked out the asic miners

Scrypt-N had ASICs?

KNC Titan had support for it (https://github.com/KnCMiner/titan/releases/tag/v1.95), though not sure it was ever "optimized".  Not sure of other manufacturers.

[djm34]

work finished since a good week... (actually I helped vtc team in tuning their new algo, so the code even predate the algo  )

Interesting. The  vertcoin botnet has over 50% of the hashrate, will they be able to hardfork?

- Botnets have free electricity
- Botnet owners will dump the coins for bitcoins and crash the market price.

well it is still in their best interest... since exchanges, pool will switch
(and it won't be the case anymore at time of the fork and they might end-up on their own fork, considering wallet will have a new check point set)

actually they will probably mine the last block and be out without noticing

ps: the situation was much worst when they forked out the asic miners

Scrypt-N had ASICs?

KNC Titan had support for it (https://github.com/KnCMiner/titan/releases/tag/v1.95), though not sure it was ever "optimized".  Not sure of other manufacturers.

optimized or not, it was pretty obvious they were there when the scrypt-n support was added

[sp_]

Submitted a tiny speedup in groestl.  (3 commits)

quark is now peaking at 16500 on the gigabyte 970 windforce factory clocks