Sorry guys but this thread is too big for me to follow. I modified the pascal cuda miner and optimized it, now it's doing 900 Mh/s on a 1070 consuming just 100W. Is it good? Worth working on it, now that there are pools?
|
|
|
When pool hashrate is much higher than network, it may be due to a high number of orphans, so be careful when you do your own math.
|
|
|
Again this is obsolete miner software. Look for a newer sgminer fork.
|
|
|
Let me make a little briefing on the coin issues. Two of them are known to me: one is blockchain sync from scratch and the other is wallet corruption. The former can be mitigated by downloading a blockchain snapshot: I will do this periodically and I've seen other nice people doing the same. The latter I could never reproduce: it must be something related to the environment the wallet runs into, or is compiled with. For this reason, I'd like to collect some information on the operating system (and library versions, when different from the default) of both the people who are experiencing the corruption issue and those who aren't.
Here is my contribution:
Corruption issue: never experienced during months of use. OS: ubuntu 16:04 Libraries: standard provided by the distro.
|
|
|
do pascal move
ESP is the next pascal Pascal coin came out with their wallet and it contains a messaging system like the one that was being worked on for ESP. pascal beat esp to release with this feature. perhaps in an initial feature release, though I have not seen pascal to be honest. Our features are different than what others offer. As some of the current ones already show. There's more than one way to bake a cake and not all ways are a good way. Thank you for clarifying that messaging has been attempted before. We just want to perfect it Features are important, but there are so many innovative coins which just die or are not solid enough ;-) ESP is growing steadily and, IMHO, is going to be a solid and long living coin.
|
|
|
yep, you need a few more posts on the forum for that, add a thanks in my cpuminer thread the coin diff should rise a bit with that There were already a few miners with that hashrate ;-)
|
|
|
How do you get the cuda miner to work? I get only crashes.
I had to fix here and there and now it works, I even improved it a bit, but solo mining with such difficulty is no fun and even the hashrate reporting is crap for any serious optimisation work. Some parts should be rewritten.
|
|
|
I will write more comments as soon as I have some time, but I wanted to say something about the opensource thing now. You can keep it closed, of course: it's your baby. But people will not trust you. And you will get much less support and help from the community. You could, eventually, keep some parts closed.
|
|
|
I asked on their thread and they replied it's in the miner but I just can't find it anywhere.
|
|
|
again, where the hell is the miner proxy jar? can anyone point me to the file or the archive containing it? thanks!
|
|
|
I'm on Linux. Can't find the proxy anywhere: cuda miner, opencl miner, Linux wallet, wallet sources, miner binaries...
|
|
|
where is the pascalcoinproxy2.jar?
|
|
|
Consider improving Veltor, there is someone in the VLT thread advertising 40Mh for the Fury for 10Bitcoin. Probably not going to sell any, but if that speed is legit similar and/or faster should be achievable with the 1070.
the slowest (by far) algo on the veltor chain is streebog, which is based on lookup tables, like groestl. amd cards have an edge on that, because of the faster shared memory (local memory, like it's called in opencl). unless someone can do a bitsliced streebog (and there are no implementations of it that I know of) don't expect the same level of performance of amd.
|
|
|
No, this algo also switch the first one which can be much slower than others, unlike x11evo.
Well some are very fast due to the possible midstate, and some much slower, like groestl
It is a permutation, so they just change order (including the first, of course). Are you sure the compile can optimize the midstate by itself?
|
|
|
I need some more cheese in the mouse trap. Why not some Fricotal?
|
|
|
Added on yiimp.ccminer.org, i like the algo name Imported also madzebra work in cpuminer-multi (as is for now, might be optimized later) Reminder: This yiimp pool doesnt allow BTC addresses... for now this yiimp algo implementation code is not released, i need to do some tests due to the variable hashrates of the different algo permutations. I accept MAC donations on MCchQ675kJYeXoCACgkKx6xakubE8CAQ5A. If some day you want to do a gpu miner, you can contact me... Different speed? Looks weird: they are the same algos, just chained with different order...
|
|
|
Thanks George5, let's hope they finally fix their wallet.
|
|
|
whaaaaaaaat wow, you serious pallas, when did you receive this,
wow this makes my year, haha..
i mean last year.
Today.
|
|
|
Latest news from Poloniex about XCN deposits:
"rest assured that we have not given up on it as new information and possible solutions keep reaching us"
|
|
|
i'm trying to mine x11 marijuana coin with that miner on suprnova but it doesnt work. i have an amd card
Unless you really know what you are doing, don't mine x11 with gpu. There are asics for that. why? Gpus are highly inefficient. You can mine ten times more coins, using the same power, with asics.
|
|
|
|