CCminer(SP-MOD) Modded GPU kernels.

[sp_]

It's now clear that you can't "just mine".
You need to have at least two of the following:

- developer skills to make your own optimisations
- very cheap or free electricity
- time to follow the market closely and choose what to mine
- daytrading skills to inflate your profit (but then you could just buy, instead of mine)

35000 gtx 970's (hashrate) have been mining etherum with success the last 6 months every day. $15 000 000 worth of coins mined in 6 months..

Etherum market cap: $ 248,616,571   
  

[sp_]

Ethereum is still hot BTW.

Bensam seems to be the only one in the thread who understand how to mine with a profit. The rest are trolls.

[sp_]

what's -U? not supported on my ethminer from repo.

-U is for cUda.
https://github.com/Genoil/cpp-ethereum/tree/master/releases
building from src works ok on Linux, on Windows it's a bit of a mess.

If you implement the block split in the cudaport like you did in the opencl version I will donate 0.1BTC.

[s7icky]

Ethereum is still hot BTW.

Bensam seems to be the only one in the thread who understand how to mine with a profit. The rest are trolls.

5% bro 5% go get it done already

[s7icky]

what's -U? not supported on my ethminer from repo.

-U is for cUda.
https://github.com/Genoil/cpp-ethereum/tree/master/releases
building from src works ok on Linux, on Windows it's a bit of a mess.

If you implement the block split in the cudaport like you did in the opencl version I will donate 0.1BTC.

then steal the source and resell im guessing?

[tbearhere]

Time to move all the compute 5.2 devices in my farm to Etherum

Ether is currently paying

Ether: 0,294BTC/GHASH   (650GHASH network hashrate (191BTC generated to the miners every day) ($72 618))
Quark:0.14BTC/GHASH

gtx 960@ether: 11MHASH (qtminer opencl)
gtx 960@quark: 11MHASH (spmod private #5)

gtx 970@ether: 21MHASH(oc) (genoils cudaminer)
gtx 970@quark: 20MHASH(oc) (spmod private #5)



I will double my profit.

@Sp  what does the 750 Ti get in hashrate ether?
Thx

[Genoil]

what's -U? not supported on my ethminer from repo.

-U is for cUda.
https://github.com/Genoil/cpp-ethereum/tree/master/releases
building from src works ok on Linux, on Windows it's a bit of a mess.

If you implement the block split in the cudaport like you did in the opencl version I will donate 0.1BTC.

What do you mean? The OpenCL dag allocation in chunks? That turned out to be a pointless exercise. It didn't improve hash rate on Nvidia and it didn't work at all on AMD. But I'm happy to implement it for you. In fact I believe I actually tried that as a remedy against the TLB trashing behavior a few months ago on the CUDA miner, but it didn't work.

[Cryptozillah]

Ethereum is still hot BTW.

Bensam seems to be the only one in the thread who understand how to mine with a profit. The rest are trolls.

You want people to keep donating and at the same time you call them trolls ? wtf

[sp_]

Ethereum is still hot BTW.

Bensam seems to be the only one in the thread who understand how to mine with a profit. The rest are trolls.

5% bro 5% go get it done already

I submitted a 5% increase in the DOOM algo @github (gtx 970 compute 5.2)
The Deep algo is also faster.

Deep is now doing 9-10MHASH on the 750ti. Same as the amd 7950 with sgminer but with 1/3 of the power.

Just for you s7icky.

[sp_]

What do you mean? The OpenCL dag allocation in chunks? That turned out to be a pointless exercise. It didn't improve hash rate on Nvidia and it didn't work at all on AMD. But I'm happy to implement it for you. In fact I believe I actually tried that as a remedy against the TLB trashing behavior a few months ago on the CUDA miner, but it didn't work.

I have problems with rigs that only have 4 gigs of memory and mixed cards. (amd/Nvidia) So the daggerfile should be loaded in chunks to avoid the usage of virtual memory.

[bathrobehero]

what's -U? not supported on my ethminer from repo.

-U is for cUda.
https://github.com/Genoil/cpp-ethereum/tree/master/releases
building from src works ok on Linux, on Windows it's a bit of a mess.

If you implement the block split in the cudaport like you did in the opencl version I will donate 0.1BTC.

What do you mean? The OpenCL dag allocation in chunks? That turned out to be a pointless exercise. It didn't improve hash rate on Nvidia and it didn't work at all on AMD. But I'm happy to implement it for you. In fact I believe I actually tried that as a remedy against the TLB trashing behavior a few months ago on the CUDA miner, but it didn't work.

You obviously know way more than I do but last time I checked that option it stated that it ignored it for GPU mining:

Creating 128 chunked buffers for the DAG
Loading single big chunk kernels because GPU doesn't care about chunks!

[djm34]

Ethereum is still hot BTW.

Bensam seems to be the only one in the thread who understand how to mine with a profit. The rest are trolls.

LOL

you made my day  

I am thinking to put that into my signature  

"Bensam seems to be the only one in the thread who understand how to mine with a profit. The rest are trolls."

(best quote ever )

[Genoil]

What do you mean? The OpenCL dag allocation in chunks? That turned out to be a pointless exercise. It didn't improve hash rate on Nvidia and it didn't work at all on AMD. But I'm happy to implement it for you. In fact I believe I actually tried that as a remedy against the TLB trashing behavior a few months ago on the CUDA miner, but it didn't work.

I have problems with rigs that only have 4 gigs of memory and mixed cards. (amd/Nvidia) So the daggerfile should be loaded in chunks to avoid the usage of virtual memory.

Each instance of ethminer keeps a copy of the full dag in system RAM to validate a solution. However, if it is not present, it will perform a slightly slower "light" validation (https://github.com/Genoil/cpp-ethereum/blob/master/libethcore/EthashAux.cpp line 267-276). If you would load the dag in chunks, send them over to the GPU and then release the system RAM DAG, it could work at a slight performance hit. However, this is not at all what I did in the opencl-chunks branch. The only thing I did there was allocating and sending over the DAG to GPU RAM in chunks.

My 1.0.3 release adds the option to load the DAG from a different file location, so you can do mixed CUDA/OpenCL mining from two instances. However this requires double the DAG RAM, so it's far from ideal. It would be much better to be able to have ethminer auto-select the right GPGPU platform per Worker thread so you can do mixed CPU/OpenCL/CUDA mining from a single ethminer instance. This shouldn't be too much of a hassle to get working. I'll see if I can get that into 1.0.5 (1.0.4 is for stratum support, which I'm fighting with currently).

You obviously know way more than I do but last time I checked that option it stated that it ignored it for GPU mining:

Creating 128 chunked buffers for the DAG
Loading single big chunk kernels because GPU doesn't care about chunks!

Yes I wrote that. The point is, that the original chunks implementation by ETH:DEV (that is disabled in the source btw), used a specific kernel that passed 4 separate pointers to each of the 4 DAG chunks into the kernel. My implementation however did load the DAG in chunks, but assumed the chunks would be allocated contiguously in GPU RAM and therefore just passed a single pointer to the beginning of the DAG into the kernel. I found this apporach somewhere on AMD's opencl forums. It turned out to work fine on Nvidia's OpenCL implementation but not on AMD's. I think there was some kind of fragmentation going on there, but as I don't own any AMD hardware, I discarded the idea of solving anything with that. The original goal of the chunks approach was to avoid the DAGpocalypse, but it seems now that the problems some people seem to have now with DAG allocation are not with the GPU RAM, but with the amount of system RAM on Windows (see here for instance: http://forum.ethereum.org/discussion/comment/18222/#Comment_18222).

[sp_]

Etherum $ 4.61 (+31.56 %)

http://coinmarketcap.com/currencies/ethereum/#charts

Market Cap
$ 354,927,928

Turnover (24h)
$ 25,871,500

Now you know what to mine...

1. Get a deposit adress from an exchange
2. Install this, it will take 5 minutes.
3. Mine.. A used r9 280x will ROI in 40 days($140 at ebay)

http://cryptomining-blog.com/6146-ethpool-now-with-new-stratum-etherum-mining-support/

[chrysophylax]

What do you mean? The OpenCL dag allocation in chunks? That turned out to be a pointless exercise. It didn't improve hash rate on Nvidia and it didn't work at all on AMD. But I'm happy to implement it for you. In fact I believe I actually tried that as a remedy against the TLB trashing behavior a few months ago on the CUDA miner, but it didn't work.

I have problems with rigs that only have 4 gigs of memory and mixed cards. (amd/Nvidia) So the daggerfile should be loaded in chunks to avoid the usage of virtual memory.

Each instance of ethminer keeps a copy of the full dag in system RAM to validate a solution. However, if it is not present, it will perform a slightly slower "light" validation (https://github.com/Genoil/cpp-ethereum/blob/master/libethcore/EthashAux.cpp line 267-276). If you would load the dag in chunks, send them over to the GPU and then release the system RAM DAG, it could work at a slight performance hit. However, this is not at all what I did in the opencl-chunks branch. The only thing I did there was allocating and sending over the DAG to GPU RAM in chunks.

My 1.0.3 release adds the option to load the DAG from a different file location, so you can do mixed CUDA/OpenCL mining from two instances. However this requires double the DAG RAM, so it's far from ideal. It would be much better to be able to have ethminer auto-select the right GPGPU platform per Worker thread so you can do mixed CPU/OpenCL/CUDA mining from a single ethminer instance. This shouldn't be too much of a hassle to get working. I'll see if I can get that into 1.0.5 (1.0.4 is for stratum support, which I'm fighting with currently).

You obviously know way more than I do but last time I checked that option it stated that it ignored it for GPU mining:

Creating 128 chunked buffers for the DAG
Loading single big chunk kernels because GPU doesn't care about chunks!

Yes I wrote that. The point is, that the original chunks implementation by ETH:DEV (that is disabled in the source btw), used a specific kernel that passed 4 separate pointers to each of the 4 DAG chunks into the kernel. My implementation however did load the DAG in chunks, but assumed the chunks would be allocated contiguously in GPU RAM and therefore just passed a single pointer to the beginning of the DAG into the kernel. I found this apporach somewhere on AMD's opencl forums. It turned out to work fine on Nvidia's OpenCL implementation but not on AMD's. I think there was some kind of fragmentation going on there, but as I don't own any AMD hardware, I discarded the idea of solving anything with that. The original goal of the chunks approach was to avoid the DAGpocalypse, but it seems now that the problems some people seem to have now with DAG allocation are not with the GPU RAM, but with the amount of system RAM on Windows (see here for instance: http://forum.ethereum.org/discussion/comment/18222/#Comment_18222).

genoil ...

i have wanted to 'play' with mining eth for a little bit now - and ive tried following your instructions for compiling in fedora 23 x64 - but it ALWAYS fails ...

i do not have a miner for eth for the nvidia cards - and when i try with sgminer ( opencl ) it too fails to compile ...

i have the latest updates for fedora 23 x64 - and cuda 7.5 ... yet - every single thing i try fails to compile the eth miner ... not many other miners fail compilation ( sp - tpruvot - djm34 ccminer ... sgminer-dev - wolf0 sgminer ) - but this one wants to be stubborn ...

how did you get it to compile on the fedora 23 that you suggested in your git wiki? ...

id really like to get at least ONE miner compiled and running - even if its just for a test - let alone mining for coin ...

any ideas? ... pm if you need ...

#crysx

[Genoil]

genoil ...

i have wanted to 'play' with mining eth for a little bit now - and ive tried following your instructions for compiling in fedora 23 x64 - but it ALWAYS fails ...

i do not have a miner for eth for the nvidia cards - and when i try with sgminer ( opencl ) it too fails to compile ...

i have the latest updates for fedora 23 x64 - and cuda 7.5 ... yet - every single thing i try fails to compile the eth miner ... not many other miners fail compilation ( sp - tpruvot - djm34 ccminer ... sgminer-dev - wolf0 sgminer ) - but this one wants to be stubborn ...

how did you get it to compile on the fedora 23 that you suggested in your git wiki? ...

id really like to get at least ONE miner compiled and running - even if its just for a test - let alone mining for coin ...

any ideas? ... pm if you need ...

#crysx

The git wiki is not mine. The only linux builds I succesfully did were on AWS G2 with Ubuntu. What is your build command line and what errors do you get?

[sp_]

Each instance of ethminer keeps a copy of the full dag in system RAM to validate a solution. However, if it is not present, it will perform a slightly slower "light" validation (https://github.com/Genoil/cpp-ethereum/blob/master/libethcore/EthashAux.cpp line 267-276). If you would load the dag in chunks,

But if you remove cpu validation, the etherminer doesn't need to use any ram to store the dag file. The dagfile is only located in the GPU memory. Should be an option to do this, and also an option to load the file to the gpu in chunks. I am too lazy to upgrade my rigs with 8 GB of ram just to mine ETHER.

[sp_]

i have the latest updates for fedora 23 x64 - and cuda 7.5 ... yet - every single thing i try fails to compile the eth miner ... not many other miners fail compilation ( sp - tpruvot - djm34 ccminer ... sgminer-dev - wolf0 sgminer ) - but this one wants to be stubborn ...

You should switch to windows. Then you don't need to compile anything. The software simply works out of the box.. Much easier, alot more profit.

[chrysophylax]

genoil ...

i have wanted to 'play' with mining eth for a little bit now - and ive tried following your instructions for compiling in fedora 23 x64 - but it ALWAYS fails ...

i do not have a miner for eth for the nvidia cards - and when i try with sgminer ( opencl ) it too fails to compile ...

i have the latest updates for fedora 23 x64 - and cuda 7.5 ... yet - every single thing i try fails to compile the eth miner ... not many other miners fail compilation ( sp - tpruvot - djm34 ccminer ... sgminer-dev - wolf0 sgminer ) - but this one wants to be stubborn ...

how did you get it to compile on the fedora 23 that you suggested in your git wiki? ...

id really like to get at least ONE miner compiled and running - even if its just for a test - let alone mining for coin ...

any ideas? ... pm if you need ...

#crysx

The git wiki is not mine. The only linux builds I succesfully did were on AWS G2 with Ubuntu. What is your build command line and what errors do you get?

i follow the the instructions as per the wiki - so there is the umbrella list that i follow ...

its a rolling compile - as the script that they produce for the compile is supposed to do it automatically ...

even when compiling separately - a couple of the components refuse to compile ...

i know im being general here - but fedora is very different to ubuntu on one main component - openssl ... ecc is the main culprit that redhat removed from the openssl implementations ... but there has been a revision on that and the secp256k ( not sure if its k1 ) of the latest openssl now has some of the curves they have omitted prior ...

when i get the chance to compile again - ill let you know - if you dont mind me sending you a pm later? ...

that would be appreciated if you can help ...

there is no ccminer implementation - and ive lost faith that sp will release one - unless its private AND windows only ... so i have to compile this on my own ...

and all the help i can get from anyone that is in the know ...

#crysx

[antantti]

genoil ...

i have wanted to 'play' with mining eth for a little bit now - and ive tried following your instructions for compiling in fedora 23 x64 - but it ALWAYS fails ...

i do not have a miner for eth for the nvidia cards - and when i try with sgminer ( opencl ) it too fails to compile ...

i have the latest updates for fedora 23 x64 - and cuda 7.5 ... yet - every single thing i try fails to compile the eth miner ... not many other miners fail compilation ( sp - tpruvot - djm34 ccminer ... sgminer-dev - wolf0 sgminer ) - but this one wants to be stubborn ...

how did you get it to compile on the fedora 23 that you suggested in your git wiki? ...

id really like to get at least ONE miner compiled and running - even if its just for a test - let alone mining for coin ...

any ideas? ... pm if you need ...

#crysx

The git wiki is not mine. The only linux builds I succesfully did were on AWS G2 with Ubuntu. What is your build command line and what errors do you get?

-OT-

Genoil, I hope you dumped my donation already. I don't want to be that guy they are telling stories about... you know, 10 000 BTC for pizza and so on...