Parallel vs Series Mining: can machines cooperate without double-work?

[roxane40]

Can ASIC mining machines mine in parallel? Can they pool their hashing power to mine in parallel and avoid doing the same work twice? Are machines in standard mining pools actually competing on one block? Do they mine in series? Or do they mine in parallel, genuinely adhering to a hierarchical tandem cooperative relationship so they don’t do double work? I mean, if I take 3 ASICs and stick them together, they’re just 3 independent machines, so yes I have 3x the chances of mining a block. But if they were to share the same mining algorithm, mining in parallel and sharing each other’s task- each mining a different part of the block so to speak, to avoid repeating each other’s calculations- they would be exponentially more effective. Is this already done? What is the software for this? Are mining pools typically series or parallel? Is this a common configuration?

[btc78]

You have a lot of questions but the simple answer here is they don’t usually do mine in series because it is highly inefficient. When you mine in series, essentially you will have to wait for the other miner to finish doing their task before you can do yours.

In a mining pool, there is a distribution of task to ensure and maximize productivity which is what mining parallel is.

[roxane40]

Right.

[DannyHamilton]

Can ASIC mining machines mine in parallel?

Yes.

Can they pool their hashing power to mine in parallel and avoid doing the same work twice?

Yes.

Are machines in standard mining pools actually competing on one block?

No. They are each working on their own separate block for every hash they compute. However, they are typically all working on creating a block for the same block height.

Do they mine in series?

No.

Or do they mine in parallel, genuinely adhering to a hierarchical tandem cooperative relationship so they don’t do double work?

Correct.

I mean, if I take 3 ASICs and stick them together, they’re just 3 independent machines, so yes I have 3x the chances of mining a block.

Correct. If they are identical ASIC with the same hashpower, running at the same clockspeed.

But if they were to share the same mining algorithm, mining in parallel and sharing each other’s task- each mining a different part of the block so to speak, to avoid repeating each other’s calculations- they would be exponentially more effective. Is this already done?

Yes.

What is the software for this?

Any standard Bitcoin mining pool software.

Are mining pools typically series or parallel?

They all work on the current block height in parallel. Block heights are handled in series (block height 1 MUST exist before ANYBODY can start on block height 2)

Is this a common configuration?

Yes.

It seems to me that you effectively asked the same question 10 times?  Why not just ask it once and then wait for the answer?

[franky1]

asics are the miners
asics dont collate the transactions, the mining pool manager does

the mining pool manager collates transactions and then adds a coinbase reward transaction to the top of the list
this is then in a merkle tree where the root of the tree is a hash contained in the header

because mining asics dont have a full node inside them they dont collate/change the normal transactions, but instead make small continuous changes to the coinbase reward transaction output(other spare/unique outputs in the coinbase tx) to create some unique modification to the merkle root hash so they can then hash the header to create a different block hash than the last change.. they do this trillions of times per second
(~200,000,000,000,000 hashes per second per asic on average(asic average speed is 200Thash per asic right now))

so a pool as a whole(manager and miners) works on the same block of transactions. yet modifying the coinbase reward so that each asic miner is doing its own random work different to each other, whilst still working on the same block of collated transactions for that block

a different mining pool collates its own transaction list which would differ from other pools. so different pools would work on a totally different transaction list thus has a different merkle root thus different hash of the header

thus miners dont work on the exact same work

mining pool managers manage the transaction list and send out the blockheader to asic machines for the machines to do the small modifications to the coinreward spare outputs. so there is cooperation to work together without working on the same work to create the same blockhash..