developer is very quiet
is there some info about SKC? trades?
Topic: [ANN] [SKC] Skeincoin 0.9.3.1 | Skein-SHA2 - page 65. (Read 161541 times)
SKC = dead.
Current diff = 5diff last week = 2
but man people with a working gpu have a crazy advantage
maybe he was referring to the developer...
Indeed.
SKC = dead.
Current diff = 5diff last week = 2
but man people with a working gpu have a crazy advantage
maybe he was referring to the developer...
SKC = dead.
Current diff = 5diff last week = 2
but man people with a working gpu have a crazy advantage
SKC = dead.
I would like to buy Skeincoin, if anyone has any for sale. Please message me.
any news from SKC developer ?
Quote
Cool but how ? Smiley did you rewrite it to AMD SDK ?
Reorder's kernel is in opencl, so it should run on anything that knows how to take advantage of the output/input of the code (not me haha).
Quote
Just hope it is an irony, IMO 'modular' is the least applicable word for it. Though it is very nice in multiple other ways of course. You won't have to "#ifdef SCRYPT" in more than one place in a true modular system.
mm It took me about four days before I realized that the precalc was overwriting whatever work I prepped. I'm certain I'm missing a hex2bin or flip, I was going to draw up the work flow. Another end is where it checks the work locally, I'm not sure the cpu code path is consistent with the gpu. I really am an amateur, and it feels like I'm fumbling around.
To get it integrated fully I'm going to have to figure out the cl state variables, and how exactly to prep the skein prestages with the work. Assuming I could do that, then even debug it, it has to wait for now. I have to work on other stuff.
Quote
I'm just wondering how you could compile the kernel for poclbm in python that you showed, to cgminer.
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice Smiley
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice Smiley
mmm as far as patching the kernel in the easy stuff has been done.
Cool but how ?
did you rewrite it to AMD SDK ? Quote
I'm just wondering how you could compile the kernel for poclbm in python that you showed, to cgminer.
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice Smiley
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice Smiley
mmm as far as patching the kernel in the easy stuff has been done. It hashes with invalid nonce hw errors at this point, not hard to do. I had the pre stages mixed up, got bent and sick of it, and passed it to redkendra. Just the last couple days I have been staying away from it.
It took me awhile to appreciate how modular cgminer really is.
Just hope it is an irony, IMO 'modular' is the least applicable word for it. Though it is very nice in multiple other ways of course. You won't have to "#ifdef SCRYPT" in more than one place in a true modular system.
Quote
I'm just wondering how you could compile the kernel for poclbm in python that you showed, to cgminer.
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice Smiley
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice Smiley
mmm as far as patching the kernel in the easy stuff has been done. It hashes with invalid nonce hw errors at this point, not hard to do. I had the pre stages mixed up, got bent and sick of it, and passed it to redkendra. Just the last couple days I have been staying away from it.
It took me awhile to appreciate how modular cgminer really is.
cgminer must be parsing the output buffer differently from poclbm (see SETFOUND macro in its kernels), otherwise looks good, at a glance. Please also note the kernel only returns diff 1 shares, for pool you are going to need to either hardcode the target or pass it to kernel.
Yes, cgminer parses it differently. I added thisCode:
// output[OUTPUT_SIZE] = output[nonce & OUTPUT_MASK] = nonce;
#define FOUND (0x0F)
#define SETFOUND(Xnonce) output[output[FOUND]++] = Xnonce
SETFOUND(nonce);
Unfortunately it never reaches this code (on Geforce GT 650) and I don't know why...
You can try a lower diff (uncomment the '& 0xc0ffffff' for example) and see what happens. You can also use printf() in the kernel for debugging if your SDK is at least OpenCL 1.1 compliant.
I'm just wondering how you could compile the kernel for poclbm in python that you showed, to cgminer.
I've looked at the others kernel in cgminer like diablo or phatk, this doesn't look the same at all.
Any explanation would be nice
thanks
Well I'm at a stage where poclbm mines but all shares are rejected so i must be doing something wrong.
I want to be sure I'm sending the correct data to your opencl kernel but unsure which parts of the state, state2 and f arrays poclbm generates i should be sending as what variable. In particular I'm referring as to what i should supply for "data16", "data17" & "data18".
Are you calculating mid state in python anywhere?
I think a couple of code snippets will make it more clear..I want to be sure I'm sending the correct data to your opencl kernel but unsure which parts of the state, state2 and f arrays poclbm generates i should be sending as what variable. In particular I'm referring as to what i should supply for "data16", "data17" & "data18".
Are you calculating mid state in python anywhere?
in Switch.decode:
Code:
binary_data = block_header.decode('hex')
job.headerX = binary_data[:76]
job.dataX = unpack('<19I', job.headerX)
in OpenCLMiner.mining_thread:
Code:
state = skeinhashmid(pack('>16I', *work.dataX[:16]))
for i in xrange(8):
self.kernel.set_arg(i, state[i * 8:i * 8 + 8])
for i in range(16, 19):
self.kernel.set_arg(i - 8, pack('>I', work.dataX[i]))
I do not use poclbm's engine to recalculate the merkle tree, just always use full getwork as it has arrived from the server. Probably it does not play well with stratum, but works to some extent.
skeinhashmid is a ctypes wrapper for skein calculation in shared lib, I believe I have posted the C function for that somewhere below the kernel itself.
Thanks reorder
I won't lie, this is still a struggle for me but I don't expect to be drip fed bits of code until it works. I'm doing this to learn and I'm certainly achieving that so thank you
Well I'm at a stage where poclbm mines but all shares are rejected so i must be doing something wrong.
I want to be sure I'm sending the correct data to your opencl kernel but unsure which parts of the state, state2 and f arrays poclbm generates i should be sending as what variable. In particular I'm referring as to what i should supply for "data16", "data17" & "data18".
Are you calculating mid state in python anywhere?
I think a couple of code snippets will make it more clear..I want to be sure I'm sending the correct data to your opencl kernel but unsure which parts of the state, state2 and f arrays poclbm generates i should be sending as what variable. In particular I'm referring as to what i should supply for "data16", "data17" & "data18".
Are you calculating mid state in python anywhere?
in Switch.decode:
Code:
binary_data = block_header.decode('hex')
job.headerX = binary_data[:76]
job.dataX = unpack('<19I', job.headerX)
in OpenCLMiner.mining_thread:
Code:
state = skeinhashmid(pack('>16I', *work.dataX[:16]))
for i in xrange(8):
self.kernel.set_arg(i, state[i * 8:i * 8 + 8])
for i in range(16, 19):
self.kernel.set_arg(i - 8, pack('>I', work.dataX[i]))
I do not use poclbm's engine to recalculate the merkle tree, just always use full getwork as it has arrived from the server. Probably it does not play well with stratum, but works to some extent.
skeinhashmid is a ctypes wrapper for skein calculation in shared lib, I believe I have posted the C function for that somewhere below the kernel itself.
Well I'm at a stage where poclbm mines but all shares are rejected so i must be doing something wrong.
I want to be sure I'm sending the correct data to your opencl kernel but unsure which parts of the state, state2 and f arrays poclbm generates i should be sending as what variable. In particular I'm referring as to what i should supply for "data16", "data17" & "data18".
Are you calculating mid state in python anywhere?
I want to be sure I'm sending the correct data to your opencl kernel but unsure which parts of the state, state2 and f arrays poclbm generates i should be sending as what variable. In particular I'm referring as to what i should supply for "data16", "data17" & "data18".
Are you calculating mid state in python anywhere?
Been fiddling with this for days now and still can't get any sense out of it. Guess my coding skills just aren't up to scratch! 
lol
lolPlease don't hesitate to ask your questions (but not about cgminer inner structure, I have spent couple hours trying to get my head around it and gave up).
I will pay .03 BTC for 30K SKC
Paying 1LTC for 12000 SKC
where can i sell my 1200 SKC?
I will pay 0.0015 BTC for 1200
Jump to: