Topic: Mining Bitcoin at higher than current difficulty (Read 104 times)

At this point I don’t believe mining is profitable.   Yes you may make some coin that can justify the costs at the current BTC price but in the long run it will fail.  The reason is that the mining market is saturated and the difficulty is set to basically make mining break even or just under even.  With the big farms the way it works is you lease a space and electricity.  In essence you are paying for the farm operator to get free energy since the rate they charge you will pay for their lease and their electricity.  So you end up with a balance of zero if you account for the miner costs, the rental+electricity since it all cancels out the BTC value you mined.   I was in this for a few years.  It’s a mugs game.

If you want to play with BTc just trade it.  You’ll never be able to beat the costs of a farm unless you have your own solar power.

Right. In terms of "normal protocols" there's only Bitcoin and everything else. A Bitcoin miner is dumb and doesn't know what network it's on or even what "normal protocols" are. It only know how to solve one particular problem, whether on mainnet, testnet, or altcoin. Yet there's more nuance to mining on testnet than just proving if code and hardware doesn't crash and can handle lower diff scenarios. For example, one could test different builds of the same code, different batches of supposedly the same hardware, not to mention identifying the cause of any stubborn latency issues. There are many moving parts under the hood. And if one is seeing unusual patterns emerging in extremely large-scale miners, which is what I'm seeing, testnet or something similar could be used to understand what's causing them. Anyway, I think I know what to do. And I appreciate the feedback.

That ^^ was my point - testnet is useless for high diff testing so only proves that fairly low diff (by today's standards) solutions are possible from new gear or code but at least testnet follows all of mainnets I/O and process requirements. Which only proves that the code & hardware does not crash and can handle lower diff scenarios. That leaves running/testing on the mainnet as the only way to ultimately prove something works.

Rolling your own private testnet that artificially forces very high diff target is another story. Should be possible but I'm unaware of anyone doing it, mainly because it would no longer be fully simulating main or test net protocols which leaves an uncertainty about it. Of course you would also be faced with the amount of time & hash power needed to produce the desired diff unless you do something to bias the code towards it but again that would entail breaking the normal protocols...

A better place to move this to would be the Technical Discussion area where the folks are far better versed in this matter.

Testnet mining has a different difficulty than mainnet, but testnet is not considered 'fudging the inputs,' and lots of miners test equipment on testnet. Moreover, SHA256 ASICS do one thing and do them well, regardless of the chain. Again, I'm not interested in testing equipment on the mainnet, nor did I mention coming up with a 'new' miner. Mainnet is not for experimentation or testing purposes. It's the real deal. What I'm looking for is essentially a really difficult testnet, because I'm testing something. I suppose I'll just roll my own private testnet and have it. Easier said than done, I'm sure, but I've done enough mining and networking I can figure it out. But please, if anyone has any suggestions or tips or has attempted something similar, I'd much prefer to learn from others. I know that practically, this is a ridiculous endeavor, but theoretically, I want to see what numbers are.

The only way to test new gear is by mining. The process does not allow for fudging the inputs to simulate the miners returned results. All you can do is log the returned difficulty over (a long) time to see that it fits probability curves for the hashing speed used.

This why every time somebody comes up with a 'new' miner or miner code it cannot be said that it is proven to work until it finds a block.

Something I've been wondering. How would one test mining equipment against a higher difficulty than the current difficulty? Either a precise difficulty value or a range of values? Say 20% or 30% or even higher than the current difficulty. Is it as simple as modifying and recompiling code and then mining it like a personal testnet? Maybe a question like this has been asked before, but I couldn't find it.

Many years ago I operated a small-scale ASIC farm, but it's been a while since I closely examined the current state of bitcoin mining and what the future may hold in terms of difficulty. I do know that mining Bitcoin is never as simple as it appears, and that the performance of large-scale mining operations in the real world is never quite as it looks on paper.

Assuming I have the requisite hashing power, electricity, and cooling, what's the best approach to test new mining equipment at a specific higher difficulty? I'm asking from the perspective of what is realistically doable and theoretically possible, without regard for profits, sustainability, or up-front costs. Basically, I want to see the numbers for myself.  

I know the immediate question is, why? why not just mine Bitcoin at the current difficulty, since it's already very high? Why not extrapolate out from how the miners are performing now? Why would you want to waste your hashing power on an unprofitable experiment? I completely understand any cynical or incredulous or dismissive responses. And yes, I do mine at the current difficulty. I'm also aware that future difficulty levels can only be guessed.

Yet I'm genuinely curious about testing mining equipment against higher difficulty levels, both realistically and theoretically. The amount of time I want to test the miners would likely be for up to several months, longer if necessary. More time mining would yield more data.

My experience mining Bitcoin is limited compared to many of the pros here, so I very much appreciate any feedback.