Since a couple of days, i cant receive payments, i dunno why, but my maximum value for automatics payments is set to "1", but i have 3btc's in my account.
Tell me your login name please, i'll check it. |
|
|
|
Update your miners, please.Long Polling is working now, so you can download latest poclbm miner and use it with DeepBit. This new feature will eliminate stale shares and your mining efficiency may rise up to ~1.7% more on average (depends on your GPU speed and luck). What is Long Polling ? Explained here.
Font size was increased in some parts of the site. Is it better or worse now ?
|
|
|
|
Does that mean that the Quadro 6000 (cost ~$3,800) would be even better still?
Quadro may be even worse because Quaros use the same GPUs as GeForce, but at lower frequency and with drivers optimized for quality instead of speed. |
|
|
|
Is long polling useful at all for solo mining?
Yes, but it's not supported by official bitcoind, so you can't use it in solo mining yet. |
|
|
|
What is Long Polling ?
The first client software with Long Polling support will be the new version of m0mchil's OpenCL miner which is expected to be released soon.
I hope that beta test of this feature will be successfull.
ok, it's out, how to enable that? use "-a 10" instead of "-a 5" ? It's enabled by default. -a parameter is not used with long polling. |
|
|
|
|
What is Long Polling ?
Usually the pool gives each miner some "job" to be done - a range of data that miner will hash one by one in search for the share. This "job" is unique for every block and changes every time when new block appears in bitcoin chain and every time when new transaction is added to possible block by the pool. If someone finds a block and after this you submit your share, calculated from old "job", it's considered stale and is wasted.
Most GPU miners need about 10-30 seconds do try all the possible hashes in this "job" and then should ask the pool for next one, but currently it happens more frequently - about every 10 seconds miner just drops it's current job and asks for next just to be sure that it's fresh and has all new data. Miners do have options to configure this number. The longer is this interval - the higher is your chance to submit stale share.
Someone can reduce their server's load by modifying the miner so it will ask for new job only when current one is finished completely. This way it requests for a new one with bigger intervals, but also with bigger chance that it's work will be futile. They also can decide to take shares not only from this round, but previous one too (which favours CPU miners and reduces reward for GPU ones). This is a nice idea and it's effect on server load is great, but there is something else that can be done for the mining to be more fair...
Long Polling is a special extension to bitcoin RPC protocol that allows pool to notify the miner about new blocks, so it can drop calculations currently in progress and immediately start new ones without loosing time and power. I expect this to cause several percent rise of miner's efficiency.
The first client software with Long Polling support will be the new version of m0mchil's OpenCL miner which is expected to be released soon.
I hope that beta test of this feature will be successfull.
Also there is a new share counter on each worker's page, tonight it will be enhanced, improved and upgraded :)
|
|
|
|
Which web-site provides info about the next expected difficulty value and date please? http://nullvoid.org/bitcoin/ is down 99% of the time. You can also use bot in #bitcoin-dev @ irc.freenode.org Very useful for calculations :) |
|
|
|
I actually thought of the idea of selling my blocks at a profit but for a different reason.
Virgin unspent blocks are more anonymous than blocks that were acquired any other way. If I were to go buy BTC from a brick and mortar Bitcoin dealer that "knows your customer" and then I go use it to buy drugs or something, there's a link. Maybe one wouldn't chase it for drugs, but for something more heinous, maybe.
A freshly generated block has absolutely no history...
Makes no sense. If you are buying blocks, the seller knows you just like if buying BTCs. |
|
|
|
My internet broke today so my miners did nothing for several hours. Would it be possible to modify poclbm so that it continues to work on the last received task when no connection is available? Usually this can be done for 60 seconds maximum, because miner's job is changed constantly, at least every 10 minutes when new block is created in network. Even when each new transaction appears in queue, you should consider it and take new job. Your work will be absolutely useless if you just continue calculating old task. |
|
|
|
The 6990 only has 3072 Stream processors compared to the 5970's 3200 stream processors. So, correct me if I'm wrong, but the 5970 is still king.
False. The 6970 has fewer SPs than the 5870 but still hashes faster. I'd expect the same from the 6990 5970 is considerably cheaper per GH/s. |
|
|
|
And no, the slower miners don't see any negative impact, because we accept shares, and pay them for shares for (CurrentBlock) and (CurrentBlock - 1). If a block is solved every 600 seconds, it would take them 1200 seconds to have their new shares for a getwork become stale. So if i understand correctly, slow miners will be subsidised from money of honest GPU users ? Just asking. |
|
|
|
m0mchil's poclbm is not supported by your pool ?
Tried to connect with jgarzik and everything is fine, but with poclbm all i got is "Problems communicating with bitcoin RPC"
Probably I apply hotfix, when you try to connect :) And recheck your miner settings. Tried different settings and your example too, but no results. Jgarzik's miner still works fine. |
|
|
|
|
m0mchil's poclbm is not supported by your pool ?
Tried to connect with jgarzik and everything is fine, but with poclbm all i got is "Problems communicating with bitcoin RPC"
|
|
|
|
if i wanted to build a mining rig, would it be better for me to buy a USED GPU? With the rapid devaluation of computer parts, i can potentially get more computing power for my dollar. has anyone considered this? would this work? are there any problems with this?
Only used, of course. If you tested it and it's working - it will work fine :) |
|
|
|
Do these numbers seem accurate for my hardware?
The system has been hashing for 17 hours and is stable.
It's not optimal to mine solo with ~70 MH/s because it's more likely that difficulty will increase before you find a block - The average time to generate a block at 70000 Khps, given current difficulty of 55590.23763914 , is 5 weeks, 4 days, 11 hours, 27 minutes, and 12 seconds (and next difficulty increase is tomorrow). So you better join a mining pool. |
|
|
|
I have a Windows XP machine with an ATI Radeon HD 2600 Pro on it. Dunno if this card is any good either but it'll probably be better than trying to mine with the CPU.
So, any help would be appreciated.
You can perfectly mine on Windows machine (with a compatible video card, of course). At least if you got no more than 4 GPUs in one PC. |
|
|
|
Pardon me if this is a stupid question, but I assume that the required efficiency excludes CPU miners, is that correct?
I believe so (effectively), from what I've read so far. But not due to efficiency; instead, due to Stales. That makes sense, thanks for making it obvious. Someone should start a CPU-only mining pool. ;) There is no need for this. CPU miners can work fine with usual pool and get MORE reward than from special CPU-only pool. |
|
|
|
We got our 1st block :D
What is the "Purity" of a block ? |
|
|
|
By the end of a round, 1 share should = 1 share, not 5% of 1 share.
So your argument is purely based on the irrational emotions of the miner as opposed to mathematical certainty of the statistics. As slush said, most people don't really understand statistics or probability, so they will search for the pool that seem "more suitable" for them. And I saved their souls :) |
|
|
|
If it's not push-based and still based on getwork, by reducing askrate, you are increasing the chance of stale solutions. With 20s askrate, you are working on stale blocks 10/600=1.7% of the time and more when the blocks are found faster than every 600s.
Actually i think that 20s askrate will give more than 1.7% stale. But i hope that they ask at least once per minute to include new transactions :) |
|
|
|
|
Show Posts
