Mining with a computer cluster. Does it make sense?

Does it make sense to try and combine the hash rate of several computers through a master node and deliver all of it as one to the cloud mining platform, or maybe it does not make sense and you get the same, or better results by letting each machine mine individually?
Did any of you guys try that, before?

Thanks

Some people claim using a proxy for delivery improves performance. Think the stratum protocol has it implemented automagically. Fairly new to mining so I could be wrong.

Hi @vprf
Find in this post how a proxy can improve mining earnings :


Basically :

  • less hashrate power = less & smaller works, so potentially more “stale share”, and then less earnings per share
  • aggregated hashrate = more & bigger works, less/no stale shares, then better earnings.

Thank you.
I’ll take a look.

Hi.
This is a very interesting implementation. Thanks for your work.
But, I think that the nodes still mine separately after receiving the instructions from the main node.
I was wondering if there is a way to throw combine the hash rate of the nodes and “throw” it at the algorithm, maybe getting a significantly bigger hash rate than any single node mining on its own.

Do you get an advantage by aggregating the nodes CPU power, and is it even possible to do so?

thanks for your thoughts.

Hi @kofif

But, I think that the nodes still mine separately after receiving the instructions from the main node.I was wondering if there is a way to throw combine the hash rate of the nodes and “throw” it at the algorithm

This is exactly what stratum proxy does: “hashrate clustering”

  • When you are mining with multiple GPUs on a worker, the pool send you a work (target + hash) and all the GPUs of this miner try to solve it.

  • Now, when you connect an other worker to the pool, the poll will send an other work (same target, other hash). So the GPUS of the first miner work on share 1 and the second GPU-set (worker 2) on share 2.

  • Now, if you use a proxy, your computers cluster (behind the proxy) is “declared” to the pool like an unique miner and all GPUs try to solve the same work (target + hash). So, if you have 5 rigs with 6 GPUs, the proxy act like an unique 30 gpus miner with aggregated hashrate.

Best

That is really cool.
I will try it when the cluster is ready and let you know how it works. :wink:

Thanks
Ilan

1 Like

Just an other point : pay attention to the pool you choose.
Use a pool allowing “shares difficulty raising” properly, like dwarfpool and nanopool.

hi …
how about the cluster mining?
any results?
did it work?

Hi.

I did not create the cluster.

The overall price of buying the 5 Rock64 boards and peripherals would be more than $700 and I could not see a way to recoup this kind of investment through mining.

Ilan

well actual NIMIQ mining with cpu is rentable…
a lot use VMs… cluster should give more power

i have 3.2MH/s atm with VMs … makes me around 31,275.43 NIM / 91.92 USD a day

and price is moving… eyes on NIMIQ

greets

Please share info! How are you mining with VM’s?

Well get you a NIMIQ wallet and connect to Sushipool
join Discordchannel
we help you
a good howto you find at NIMIQ fun pool
but sushi has better scrips to mine

https://sushipool.com/#!/
https://safe.nimiq.com … wallet
Latoken… exchange
Tradesatoshi… Exchange
MS azure and Google cloud to get VM s

People BOINC like this too

here you go with my current hash and income…
Hash Rate: 7.2 MH/S
Geräte: 90 / 91
Wallet-Guthaben: 8527.412 NIM / 18.129 USD
Alltime <income: 404544.593 NIM / 860.062 USD
Pool-income: 6322.452 NIM / 13.442 USD
income day / Tag: 77,638.21 NIM / 165.06 USD
income month / Monat: 2,329,146.40 NIM / 4,951.77 USD

so you can see this is quite profitable
https://bitcointalk.org/index.php?topic=4842615.0

1 Like