barwizi (OP)
Legendary
Offline
Activity: 882
Merit: 1000
|
|
May 13, 2013, 08:29:46 PM |
|
|
|
|
|
ohhaithere
Newbie
Offline
Activity: 56
Merit: 0
|
|
May 13, 2013, 08:32:41 PM |
|
I bet it's already been done, why would anyone want to release that? 5000YAC is a joke of bounty compared to what they could be pulling in every day. It should be trivial to accomplish, I'd have done it ages ago if I didn't have a bunch of final exams next week to study for.
|
|
|
|
barwizi (OP)
Legendary
Offline
Activity: 882
Merit: 1000
|
|
May 13, 2013, 08:36:16 PM |
|
I bet it's already been done, why would anyone want to release that? 5000YAC is a joke of bounty compared to what they could be pulling in every day. It should be trivial to accomplish, I'd have done it ages ago if I didn't have a bunch of final exams next week to study for.
My skills are lacking in the coding section, and i am also not a rich man. ergo i made it a community request. Though if you really were capable of such a feat , i believe you would have done so already due to what you would "rake in".
|
|
|
|
Hydroponica
Full Member
Offline
Activity: 182
Merit: 100
fml
|
|
May 13, 2013, 09:05:19 PM |
|
You should probably offer a coin that's actually worth something Specially since if this became available to the general public, the whole premise of YAC would be defeated, and it would become even more worthless
|
|
|
|
barwizi (OP)
Legendary
Offline
Activity: 882
Merit: 1000
|
|
May 13, 2013, 09:10:22 PM |
|
You should probably offer a coin that's actually worth something Specially since if this became available to the general public, the whole premise of YAC would be defeated, and it would become even more worthless if i had more of any other coin i'd gladly pay to get first crack at it
|
|
|
|
monokaskade
|
|
May 13, 2013, 09:13:41 PM |
|
Its not so trivial if you don't have any experience coding OpenCL kernels
|
|
|
|
WindMaster
|
|
May 13, 2013, 09:33:09 PM Last edit: May 13, 2013, 10:32:18 PM by WindMaster |
|
Its not so trivial if you don't have any experience coding OpenCL kernels
Probably people without OpenCL development experience aren't going to be "in the running" for a bounty of this type. The bounty amount is a bit of a joke, considering the actual number of hours needed to port scrypt+chacha/keccak(N,1,1) to the cgminer kernel and the resulting profitability using it for mining rather than releasing it. Now, make it a 50 BTC bounty and we'll just see what springs forth..
|
|
|
|
WindMaster
|
|
May 13, 2013, 09:35:02 PM |
|
Is that a bounty (of whatever amount, 5000 YAC is just too low) to whoever implements it in a provable way, or actually releases it?
|
|
|
|
TheJuice
|
|
May 13, 2013, 10:10:18 PM |
|
Pointed here by my buddy. I could do this; but at .0006 BTC per YAC.... is it really worth it?
|
|
|
|
turtle83
|
|
May 13, 2013, 10:26:11 PM |
|
The problem here is i dont think that OpenCL code for scrypt-jane exists out there. IIRC the kernels for sha256 and litecoin type scrypt was based on pre-existing code developed for other purposes then adapted for mining...
So if someone were to write from scratch, maybe have some influence from litecoin-scrypt, id assume it easily taking up 10 - 20 hours minimum for an experienced OpenCL developer.
1) Now if said developer is into mining, 5k YAC is pocketchange compared to what they get by not sharing. 2) If said developer is not a miner, then ud need to pay a lot more... maybe like 10 - 20 BTC easily. Thats assuming their time to be worth 1 BTC/hour.
So, find person that fits category 1 and cares more about community than personal goal... Then theyd not care much about the bounty anyways.
Having said that i can add in 2k YAC to the cause, to be in your "donators" category... i am not that proficient in C/C++, but i can help with testing early code.
This project needs to be split into 2 tasks:-
1) The cgminer side of things. The side that validates the nonce returned by the kernel. This should be fairly easy, given that the c code for it exists. 2) The OpenCL kernel... this would probably be a lot more complec. Way over my pay grade. Would need some sample test cases to help with development, especially if the developer is non-miner.
|
|
|
|
WindMaster
|
|
May 13, 2013, 10:31:33 PM |
|
BTW - scrypt-jane isn't a hashing algorithm. It's a software library that supports several different scrypt-based hashing algorithms, including scrypt+Salsa20/8 as used in Litecoin and scrypt+ChaCha20/8 as used in Yacoin. The library is here: https://github.com/floodyberry/scrypt-janeAssuming the OP wants cgminer for Yacoin specifically, a change in the title of the thread is probably warranted, as "scrypt-jane" doesn't refer to the hashing algorithm in Yacoin. It's ambiguous which scrypt-derived algorithm one is referring to if you say "scrypt-jane".
|
|
|
|
TheSwede75
|
|
May 13, 2013, 10:32:02 PM |
|
There is already GPU miners running Yak. It's just such a derpy diff that it's not worth it even running GPU mining when you can just throw together a AWS instance and 51% it for a few $1.000.
|
|
|
|
ymer
|
|
May 13, 2013, 10:33:55 PM |
|
There is already GPU miners running Yak. It's just such a derpy diff that it's not worth it even running GPU mining when you can just throw together a AWS instance and 51% it for a few $1.000. Do you have any ideas about the difference from CPU to GPU?
|
|
|
|
WindMaster
|
|
May 13, 2013, 10:37:53 PM |
|
Do you have any ideas about the difference from CPU to GPU?
The difference is significant. Unless you're using a 4-processor Xeon E5-2670 server as the point of comparison.
|
|
|
|
relm9
|
|
May 13, 2013, 10:39:16 PM |
|
There is already GPU miners running Yak. It's just such a derpy diff that it's not worth it even running GPU mining when you can just throw together a AWS instance and 51% it for a few $1.000. With AWS instances, it would take way more than a few grand to launch a 51% attack.
|
|
|
|
turtle83
|
|
May 13, 2013, 10:45:26 PM |
|
BTW - scrypt-jane isn't a hashing algorithm. It's a software library that supports several different scrypt-based hashing algorithms, including scrypt+Salsa20/8 as used in Litecoin and scrypt+ChaCha20/8 as used in Yacoin. The library is here: https://github.com/floodyberry/scrypt-janeAssuming the OP wants cgminer for Yacoin specifically, a change in the title of the thread is probably warranted, as "scrypt-jane" doesn't refer to the hashing algorithm in Yacoin. It's ambiguous which scrypt-derived algorithm one is referring to if you say "scrypt-jane". Interesting, i assumed scrypt-jane was the algo. </n00bmode> https://github.com/ckolivas/cgminer/blob/master/scrypt130511.cl#L712By the way is Yacoin mining with OpenCL almost as simple as swaping out salsa with chacha20?
|
|
|
|
ohhaithere
Newbie
Offline
Activity: 56
Merit: 0
|
|
May 13, 2013, 10:49:47 PM |
|
BTW - scrypt-jane isn't a hashing algorithm. It's a software library that supports several different scrypt-based hashing algorithms, including scrypt+Salsa20/8 as used in Litecoin and scrypt+ChaCha20/8 as used in Yacoin. The library is here: https://github.com/floodyberry/scrypt-janeAssuming the OP wants cgminer for Yacoin specifically, a change in the title of the thread is probably warranted, as "scrypt-jane" doesn't refer to the hashing algorithm in Yacoin. It's ambiguous which scrypt-derived algorithm one is referring to if you say "scrypt-jane". Interesting, i assumed scrypt-jane was the algo. </n00bmode> https://github.com/ckolivas/cgminer/blob/master/scrypt130511.cl#L712By the way is Yacoin mining with OpenCL almost as simple as swaping out salsa with chacha20? That's what I was thinking, and also a SHA2 hash needs to be changed to SHA3 i believe. The code in that scrypt cl file is tedious to write, but technically trivial code. It's just mathematical operators to follow the algorithms.
|
|
|
|
turtle83
|
|
May 13, 2013, 10:54:16 PM |
|
BTW - scrypt-jane isn't a hashing algorithm. It's a software library that supports several different scrypt-based hashing algorithms, including scrypt+Salsa20/8 as used in Litecoin and scrypt+ChaCha20/8 as used in Yacoin. The library is here: https://github.com/floodyberry/scrypt-janeAssuming the OP wants cgminer for Yacoin specifically, a change in the title of the thread is probably warranted, as "scrypt-jane" doesn't refer to the hashing algorithm in Yacoin. It's ambiguous which scrypt-derived algorithm one is referring to if you say "scrypt-jane". Interesting, i assumed scrypt-jane was the algo. </n00bmode> https://github.com/ckolivas/cgminer/blob/master/scrypt130511.cl#L712By the way is Yacoin mining with OpenCL almost as simple as swaping out salsa with chacha20? That's what I was thinking, and also a SHA2 hash needs to be changed to SHA3 i believe. The code in that scrypt cl file is tedious to write, but technically trivial code. It's just mathematical operators to follow the algorithms. But where is the "N" going in scrypt130511.cl ? I know for litecoin its something static. I see this many times in scrypt_core Is 1024 the N of litecoin and thats being hardcoded in there?
|
|
|
|
WindMaster
|
|
May 13, 2013, 11:00:58 PM |
|
But where is the "N" going in scrypt130511.cl ? I know for litecoin its something static. I see this many times in scrypt_core Is 1024 the N of litecoin and thats being hardcoded in there? Yes, N is hard-coded in the OpenCL source for scrypt(1024,1,1) for Litecoin. N changes so slowly for Yacoin that you might as well hard-code it as well. Or #define N to be 32 at the top of the OpenCL file so you can change it to 64 tomorrow with a single line tweak when N=64. In the long run, to automate this, I'd just modify the cgminer source to read in the .cl file, modify the #define line with the appropriate value of N, write it back out to another .cl file, and use that .cl file as the source for compiling the OpenCL program. Remember to change the hashing algorithm in the main cgminer source first, since it needs to check the hashes. If you get it working correctly with cgminer doing CPU mining, and the shares are being accepted by, say, pushpool (or the YAC-modified pushpool over at the yac.dontmine.me pool), you can then move on to changing the OpenCL kernel. Or solo-mine Yacoin testnet, or make your own genesis block for Yacoin and solo mine off-network with very low difficulty so you can solve blocks quickly. That way you already have a known-good implementation for the hashing algorithm to check the hashes coming from the GPU.
|
|
|
|
efx
|
|
May 13, 2013, 11:03:58 PM |
|
N changes so slowly for Yacoin that you might as well hard-code it as well.
|
|
|
|
|