 Show Posts
|
|
Pages: [1] 2 3 4 5 6 »
|
Is there any comprehensive guide for selecting 'rawintensity' and 'worksize' params? What I have found is people trying to 'hit the sky' and exhange each other with 'good' (proven) values.
worksize - it is ocl localworksize of your workgroup. 64 is kind of optimal for amd on most algos(not for equihash i think, this algo is more complex than just compute hash). For amd it will vary from 1 to 256. Best sizes if it is divisible by 64. Or at least 2. But again it depends from algo and its implementation. rawintensity - when miner compute hashes, it check nonce from 0x00000000 to 0xFFFFFFFF. You can't say to gpu: compute this whole range at once, that wouldn't be effective. So you should split this range in pieces(in this case it would be global worksize in opencl terms. You can think about it like a number of hashes that sgminer would compute before verifying gpu's work on cpu and moving on to a next piece of nonces). rawintensity would be the size of such piece. It can be selected based on hardware properties of your videocard, or to be chosen experimentally(this is what most miners do, lol). xIntensity or intensity arguments allow you to set this rawintensity without computing it by yourself, as i remember intensity will set globalworksize to 2^intensity. I don't know if you can apply this to equihash impl in sgminer, i'm talking based on exp with other algos. I think Zawawa can tell more, or someone who ported his impl into sgminer-gm. UPD: actually, here is your guide: https://github.com/genesismining/sgminer-gm/blob/master/doc/MINING.mdwith intensity and worksize description. |
|
|
|
gtx 1070 (366MHASH) 3x gtx 970 (270-300 MHASH) 100% cpu load. Unusable.  About 100% cpu load: say thx to nvidia that they don't give a shit about opencl support and bug fixes. |
|
|
|
I have an important question for you all!
Would you guys support a coinswap?
Basically, I made mistakes, I get to feel these mistakes now and would jump past them by switching to a completely new coinbase.
Ready for segwit to be implemented and move to LN.
This would cut development by many many weeks/months and we'd get to our goal very faster.
If you oppose a coinswap, please don't just say "No" but explain with arguments why you think its bad.
J.V.
In this case ppl should be widely informed about that. To have time to prepare. I'm ok with it. |
|
|
|
Balance is not confirmed - all right?  It takes 2 days, 2 hours to confirm I think you should add big red label on veltor pool about it. Otherwise you will answer to the same question every day  |
|
|
|
I had 23.5mh/s in sgminer
How many graphic cards 380х do you use for 23.5 mh/s? 1 |
|
|
|
Well. ocminer, your hash calculation on pool was perfect. I had 23.5mh/s in sgminer and 20-25 at pool, normal hr fluctuation. After your first changes it was smth like 40mh/s on pool. Now it is swinging: 25-33-36mh/s
did you say Sgminer? which gpu? 380X@1100/1500 |
|
|
|
|
Well. ocminer, your hash calculation on pool was perfect. I had 23.5mh/s in sgminer and 20-25 at pool, normal hr fluctuation. After your first changes it was smth like 40mh/s on pool. Now it is swinging: 25-33-36mh/s
|
|
|
|
Most people don't understand that mining is a profession. The others can buy.
Working as a dentist is a profession too, that doesn't mean a dentist can make his own equipment. Yep. He pays for it. And mining software is an equipment too. |
|
|
|
|
Yep. Windows build with cuda 8, speed is down in a little bit. at least x86 build.
|
|
|
|
i can install vs2015 and try to compile, but i don't remeber now all the things needed, aside from cuda 8 and vs2015 what else is needed? all the dependencies i mean
to compile with cuda 7.5 or older you need vs2013. vs2015 isn't supported. Don't know for cuda 8. i remember i need something else besides those two things, but maybe i'm wrong, last time i compiled something was in 2014... some google search lead me to, vs2015 being supported only up to version 1 and with cuda 8 only, version 2 and 3 are not supported mpir, jansson, pthreads, curl, openssl. For this fork. in sp's fork everything in repo already. |
|
|
|
yeah sp is mia, he said he don't want to continue this, you should make your own thread Pallas, from what i understood you are a coder also, or maybe you have one already
No, I don't have a thread for "ccminer", even though I worked on it for more than a year now. I've submitted some improvements and made some private work. I may fork sp's ccminer and continue from there BUT: - the code base is outdated, we should start from latest tpruvot's; but then it would make more sense to submit to his own fork directly (I did that too). - sp_ said he might eventually post his optimisations on a newer codebase so I didn't want to do a duplicate work. - applying all sp_'s optimisations to tpruvot's is no easy nor quick task; it might take months. - some of the optimisations are on dead or almost dead algos; it might be better to restart from scratch... don't know. last time i checked some tpruvot algo were not running at the same optimized hash of sp last version(not talking about private) like qubit and akin i think it's better to fork from sp(so you get all the boost already), but if he will do it later at some point, then maybe we can just wait, but not sure his last message leave much to be desired... Hell no. I ported pool management from trupvot's in my private fork. And that wasn't an easy job. From the other side: no matter what you will choose as main codebase, this will take a lot of time in both cases. Just in my opinion: port from sp to trupvot's will be easier. |
|
|
|
I don't care for how long it has been here. There is a link to github in a first post and only gitignore presented.
No one cares about what you care about. Not sure that is going to win me over. I still have no answer to my original question. Cheers. This is probably my last post in this thread. Okay, which one is the sockpuppet? This account is certainly a sock puppet. But it has nothing to do with Elon. My original question - Is poker on the blockchain or on a centralized server? I have burned a legendary account and will burn this one. Both for kicks. Sorry, I thought you were trolling about the dev not working for free. This is centralized poker. You should take a look at the whitepaper. I think if there ever was something that could make decentralized real time poker a reality it might be based on what the dev proposes. But I don't know if he realizes it. There's no discussion of poker in the WP at all. 1. nobody is talking about coding for free, if u will check my posts: i against giving code for nothing. 2. there is no mention in first post about open/closed sources. And in same time link to github with gitginore only(lol). In case if src is closed, i hope dev at least gave them for review to exchange owners. Nobody in health mind will install some unknown shit on his server. |
|
|
|
Lol, no source code on github, but windows and mac binaries from somewhere? Are u fucking kidding?
Do some research. This project has been around for 2 years. Practically everyone posting favorably about it are heroes because we've been following this project for so long. The lead dev has been around forever and is even trusted by Bittrex outright to certify projects for addition. This project was in testnet for about 5 months. The worst anyone has come up with on this dev is that he won't work for free indefinitely. That tells me he is at least not crazy. I'll be excited to see the real professional trolls show up. You guys are too easy. I don't care for how long it has been here. There is a link to github in a first post and only gitignore presented. |
|
|
|
|
Lol, no source code on github, but windows and mac binaries from somewhere? Are u fucking kidding?
|
|
|
|
This coin/algo is property of Nvidia miner atm  could huge improvement be done on optimizing AMD miner i know not yep, and not even a tip for my work ^Main reason why normal amd miner is not in public. I wonder why coders for nvidia still giveaway cookies for free. |
|
|
|
Can someone do a better optimization for AMD users? Two r9 380 4g give 3,200 mh /s avg. and only one gtx 750Ti provides the same mh/s.
My 290 make 3,000MH/s if im lucky  Better mine ETH and dump for this for amd miner i see at http://www.mininghwcomparison.com/list/index.php?brand=amdsome peopke running R9 390 and get 15mhs X11evo and X11 is same ? any calculator how much you can get with 3mhs? No, it's not the same. Hell no. Amount of changes that should be done is pretty small. You don't even need to recreate a bin every day. It is even easier to write it in a such way. |
|
|
|
so what are you trying to said to me,this is an paid job ? And nvidia miner,who have paid for that ? Mount it up and post your xre/btc adress,people will throw No, from my 2years exp it is not worth to throw work at public. My point is: -It is not hard to do and there is a high probability that devs have private miner for themselves. If they don't, well they have a pretty big problem. -No, there is no nvidia advantage for this algo. Only diff between amd and nvidia here, that tpruvot was a honest one and released correct code without writing shit as coin devs did. ppl, stop talking about things that you do not understand. Read x11evo code first. |
|
|
|
Current Difficulty 50.5386
Net Hash: 2,178.88 MH/s
holy crap LOL this is next darkcoin, its begun true Also Im waiting update from dev to get AMD card better performance(at least 300-400% more performance lol) it takes ~2hrs with debugging and testing to implement it in a slightly dirty, but still more productive way. Another 1hr to get rid from bin recreating completely. But not worth it, almost no speed diff with a dirty way. |
|
|
|
Hello all,
Tried all versions of sgminer on different pools and always get this error:
[23:22:32] C:\Users\KOLJOM~1\AppData\Local\Temp\\OCL7132T5.cl:59:10: fatal error
: 'blake.cl' file not found
#include "blake.cl"
^
1 error generated.
error: Clang front-end compilation failed!
Frontend phase failed compilation.
Error: Compiling CL to IR
[23:22:32] Failed to init GPU thread 0, disabling device 0
[23:22:32] Restarting the GPU from the menu will not fix this.
[23:22:32] Re-check your configuration and try restarting.
Press enter to continue:
GPU is MSI 380 4G
Driver 16.5.3
Latest AMD driver have an error in OpenCL compiler and can't find files to include. Easiest solution just to revert drivers, i use 15.11, works best for me. BTW, for eveyrone who is saying: best algo, low temps. Are you all drunk? It is low temp just because miner written in a shitty way. |
|
|
|
|
