Avalon added intelligent auto overclocking/undervolting in the Avalon4 so you might be seeing the initial settings vs what the miner may have headed towards on its own. I have no idea on the time periods of adjustments but I'd assume they're measured in hours rather than minutes.
Yes I've experimented with the code that speeds it up over time. The device needs to have heat soaked and be relatively thermostable otherwise they have to be ramped up very slowly. Dealing with 16nm is much tricker in this regard. The original code took 2 hours to stabilise. I've experimented with changes that reach stable levels in 2 mins. |
|
|
|
|
After running mine for a day, it stabilised at 6.3Th for 1040W at the wall, though I'm using gold PSUs so expect a bit less power usage with platinum.
The more interesting thing is when I undervolted it to 4700mV, it seemed quite stable at 770W for about 10% less hashrate. Lower levels of voltage seemed to kill the hashrate or were unstable. I'll conduct more experiments in a couple of days when I'm not in offpeak electricity to see exactly what that undervolted hashrate is.
|
|
|
|
Going to bounce the pools once more this morning with further restarts while another minor upgrade is added along with segwit signalling.
Restarts complete. Mine on! |
|
|
|
Best share at 382969404792, orphan?
Edit: can't find it on the orphan list. Something wrong with pool and didn't submit the block?
Looks like it was a stale share. The pool will try and submit stale shares anyway since it has nothing to lose, but it was rejected by bitcoind since it had already registered a new block by then. |
|
|
|
|
Locking this topic. Enough answers. You cannot meaningfully mine bitcoin with a normal PC and you won't earn even 1 cent per day mining with it.
|
|
|
|
I bought 50 more S9s they should come within 20-25 days.
If I connect them all plus my 3 current S9s would I affect this pool.?
Yes you will affect this pool. You will increase its chances of finding a block proportionate to the extra hashrate  |
|
|
|
Here you can find most profitable pool: bitcoinchain.com/pools
these pools are by # of blocks found and have nothing to do with profitability.
They are mostly Chinese pools so if you are mining from other countries I would not use largest by size. I am a newbie but this list does not make sense to me.
Just because a pool finds the most blocks doesn't mean it is the most profitable pool. There are many other factors that go into determining what pool is the most profitable for individual miners. My research shows that the Kano/CK pool is the most profitable for it's members. Do your own research. Good luck and best wishes! Ah I think he means how much profit THE POOL IS MAKING which has absolutely nothing to do with the profit THE MINERS ARE MAKING. |
|
|
|
As far as I can see that site says absolutely nothing about profit, only plotting pool size which has absolutely no correlation with profit. |
|
|
|
Will this pool be signaling for segwit at any point?
I ran a search to find Kano's opinions on it and it seems he's opposed to it for a few reasons. Attempted dev control of bitcoin and disk space issues stand out the most in my memory. Any further clarifications would be appreciated. I don't visit the forums that much, so I apologise if I missed a post regarding it.
Would also be very interested to hear ck's opinion as well. Thanks.
I'm not against it. Kano and I differ in opinion on this one and ultimately this is his pool; I'm part of the administration and tech support but it's not my pool. |
|
|
|
I'm afraid this might not have been a spam attack (whatever you might mean by this). Yesterday, there were quite a few blocks that had a lot free space in them (note, I didn't consider empty blocks), and this is in striking contrast with what miners are doing today and what they had been doing just before yesterday. Whoever might want to strongly disagree, may also want to check the blockchain for themselves... So, it looks more like some miner attempting at getting more fees than someone spamming the network Well someone from the mining world has to clear this up so I'll explain. What you are seeing from miners is a combination of clunky optimisations and bad configuration; it is not completely out of malice. Believe it or not, miners often do things without knowing they're doing them and are uninformed. There are two major patterns to the blocks that are not full sized Don't worry, I read the rest of your post and just saving screen space In this way, miners are deliberately (and some part of them maybe through ignorance) pursuing profits in the form of the mining reward only, not caring much about fees collected and transactions included, thus the empty blocks. So far so good. But this doesn't explain why some blocks are only half full (and I don't mean those 750k blocks). For example, this block has only 82 transactions. A few other blocks have a few hundred transactions in them, but they are still less than half full ( this and this blocks). And such blocks were seen only yesterday, there were no such blocks today or before yesterday. What's the catch really? Pretty sure I explained - dodgy optimisation choices in a hacked coin daemon. It was notably less than 2 minutes after the previous block. The mempool would have clearly been full in a regular bitcoind at the time so there's no way a call to create a block template would have only given them a 33kB block. EDIT: Look at the block sizes here for the last 30 days and sort by average size: http://data.bitcoinity.org/bitcoin/blocksize/30d?t=lYou'll see that famously the smallest average block size is by Eligius which is the pool associated with Luke-jr... So that one's clearly not a conspiracy by the people pushing for bigger blocks. 3 of the top 4 average sizes are from ckpool users. |
|
|
|
I'm afraid this might not have been a spam attack (whatever you might mean by this). Yesterday, there were quite a few blocks that had a lot free space in them (note, I didn't consider empty blocks), and this is in striking contrast with what miners are doing today and what they had been doing just before yesterday. Whoever might want to strongly disagree, may also want to check the blockchain for themselves... So, it looks more like some miner attempting at getting more fees than someone spamming the network Well someone from the mining world has to clear this up so I'll explain. What you are seeing from miners is a combination of clunky optimisations and bad configuration; it is not completely out of malice. Believe it or not, miners often do things without knowing they're doing them and are uninformed. There are two major patterns to the blocks that are not full sized. One is the "empty blocks" where there is only the generation transaction done as SPV (light wallet) mining without full validation of the block - these are done by many Chinese pools as a speed optimisation where they mine off the header of the previously found block temporarily from another pool until they can finish the validation completely and eventually switch to a fully validated block. They claim it is a way to save themselves from the delay inherent in the great firewall of China, though it is purely an optimisation to work around their less-scalable choice of pool software and setups. Most pools outside China do not use this. Some pooled mining software has been hacked to mine empty blocks even without use of header-only mining to speed up getting out block changes to its miners as well - there are other options though that are fast on block changes without this cludge but you can't force people to choose what software they run and how they run it. See this for more discussion of empty blocks: https://bitcointalk.org/index.php?topic=1085800.0Second is purely a badly configured bitcoind either because they left it at the default or they configured it smaller by choice thinking it will provide them with a speed up to minimise their risk of orphans. The default in bitcoind is actually set to 750k and you can see many smaller "pools" or entities mining blocks of that size. Yes it's true, there are entities with millions of dollars worth of mining gear that don't know how to configure bitcoind. Additionally some pooled miners chose to actually set it lower simply to speed up work generation - notably p2pool users. Yes the blocks would ALL be full if the miners configured their set ups correctly. The reality is these choices do speed up block generation and block propagation slightly and do decrease the risk of orphans, but with each next version of bitcoind these get smaller and smaller, and I maintain pooled mining software that is designed to be quick for fully validated and full sized block but of course I can't force everyone to use my software (even though it's free.) Remember that fees are the long term incentive for miners to mine transactions into their block, but since the bulk of the fees on each block can be obtained by miners without actually filling the blocks, there is actually no major incentive for them to ensure each block is full unless they care about bitcoin transactions at large. Many miners DO care and choose their pools accordingly, but there are massive farms/entities that do not care or aren't even informed. They think they just need to get as much profit as quickly as possible from mining since the margins are now slim, and they are not aware of how their choices are affecting the network and bitcoin health. That doesn't mean their choices are actually leading to better profits by the way. |
|
|
|
|
I'm pleased to say that both Kano and I have received an A7 each from Canaan for continued support of these devices in the official cgminer, and the code has already been merged into the master cgminer.
|
|
|
|
I try to use it to mine a scrypt with a MSI GT 710, but it says plug one (no usb detected)
anyone kind enough to figure out what to do?
The official cgminer only mines bitcoin for the last 4 years. Seek help in the altcoin mining section for unofficial forks of cgminer that still mine other coins. |
|
|
|
|
I'm pleased to announce that I've received an ava7 gratis as well from canaan for continued support of this device in cgminer and am already running it on the master cgminer code on this pool.
|
|
|
|
|
I'm pleased to announce that both Kano and I have received gratis A7 units from canaan for continued support of these new devices in the official cgminer branch.
|
|
|
|
My s5s wont connect, I have tried re entering the info and power cycling. They report the pool as dead.....
same here.. dead. Pool was fine, gateway was unhappy about restart but it's fixed now. |
|
|
|
Notification: Both pool nodes will be restarted later today in approximately 2 hours in order to be prepared for the segwit activation (if it occurs) and to bring the code up to date with the latest pool changes. Miners will likely just reconnect without major interruption.
For the time being this pool is running segwit compatible block generation but is NOT currently signalling for segwit. Having said that, the pool is so small that it's unlikely to contribute to the activation decision.
Both pools updated and restarted uneventfully. Mine on! |
|
|
|
p2pool needs a high power PC. Running it on RPi would be real bad.
Concur. CPU performance aside, the RPi does not have the raw I/O performance necessary to cope with reading/writing both the blockchain and the sharechain. The getwork latency would be astronomical. 4 cores in the pi 3, as long as you 'r E not stuffing phs down its throat it should be fine. Still useless. Extra cores, but all of them run slow as shit. Per CPU core performance is everything for p2pool, especially since it's python which is mostly single threaded. |
|
|
|
Westhash rentals are dead... Is it a problem on their end or here?
Pool's fine. |
|
|
|
|
Show Posts
