Hacker News Reader

Comment by ec109685 | Hacker News Reader

This seems like a temporary problem given the difficulty adjusts? https://news.bitcoin.com/bitcoins-sinking-price-pushes-hashr...

View on HN

The used market is flooded with ASICS and GPUs so it seems the adjustment isn't working well.

Also, many large miners didn't sell their coins and instead they borrowed real money to expand and fund their new mining operations. As the interest rates are climbing and crypto prices are going down, that's not possible any more.

Nobody has used a GPU to mine Bitcoin in a decade now.

GPUs are used for eth. There is clearly a broad downturn in the mining sector.

Ethereum is mostly mined using custom hardware, it's not graphics cards really, certainly nothing you could re-sell to anybody who expects to play Bioshock or whatever. Very small operations are people buying and using actual consumer GPUs, but they tend to be absolutely destroyed by the process and then sold to unwitting buyers. I've sorted through literal milk crates of GPUs before to find a single one which could reliably display a video output after they had been used in a mining operation.

> Ethereum is mostly mined using custom hardware

Very large ETH miner here, we use GPUs. Mostly 4-5 year old tech too. Unlike BTC, ETH does not require the latest hardware because the algo, ethash, is memory bound. This is a great explainer [1].

You may be implying that the majority of the network is custom ASICs. I doubt it is more than about 30%, if that. Come august, all the ones with 5gb of ram will also disappear off the network because of DAG growth.

GPUs are also not destroyed by using them. I've got GPUs sitting in the worst possible conditions, running just fine for many years now. We have started to change the thermal paste on the hot ones though... it dries out. I don't know what you were dealing with, but that sounds like something else caused the failures as the chips themselves don't wear out like that.

[1] https://www.vijaypradeep.com/blog/2017-04-28-ethereums-memor...

It does absolutely destroy the hardware through, even very early mining farms had the sole issue of the fans in the cards burning through their bearings. A local electronics recycler had an entire crate of burned out ATX PSUs from some local mining operation, every single one of them had the fan melted off the hub. It's worth being aware that all silicon chips are temporary anyway, they all die to electron migration eventually.

I don't have fans on my cards. PSUs are generally shit (hand soldered) and that's not what we are talking about here.

do most miners choose to run cards without fans because of known issues?

The smart ones do.

Fans are moving parts and require electricity.

We also put 12 cards in a case instead of the usual 8 or 9.

This means fewer other parts too... psus, mobos, ram, ssd, case fans... all of which use power.

That is not true. I have 2 1060s and 2 580RX running 24/7 since 2 years. I recently used one to compensate for my broken gaming GPU. Everything works no issues so far. Guess it always depends but it is not like 2 years of mining will render a GPU useless.

You're within the "small miners" part of my comment.

You wrote they get "absolutely destroyed by the process". I do not understand, why you wrote this. It is not true. The stress on the gpu might even be higher while gaming, depending how the gpu is tuned.

I agree with you, he's totally wrong.

In fact, smart miners overclock/undervolt their gpus... so they are running faster, but with less power. Power is what creates heat and heat is what hurts things. Except it really doesn't.

Fans certainly die out though... we go through a lot of those on our cases. PSUs are also generally failure prone since they tend to be hand soldered on the insides.

> using actual consumer GPUs, but they tend to be absolutely destroyed by the process and then sold to unwitting buyers

This is such an urban legend. There's been quite a few tech people testing this and there's countless articles on the topic. In short - I have not seen a single one which actually found real significant issues with the hardware, even if some are listed in theory. If this was common at least one of them would buy a severely degraded batch, or we'd see some proven report about this.

Not in bulk, no.

But I'm sure there's at least a few people playing around with it on GPUs.

That's like saying people mine BTC with pencil and paper [1].

[1] http://www.righto.com/2014/09/mining-bitcoin-with-pencil-and...

Not at all. It's people who want to play around with the technology, but can't afford FPGAs or ASICS to do the job. Or maybe they could afford FPGAs or ASICS, but they want to be able to see more of what's going on behind the curtain.

Not a huge number of people, no. But then a lot of services on the Internet have a long tail.

But from what you describe, it's an open question whether anyone has used a GPU to mine bitcoin in the last ten years, as opposed to just setting a GPU to work on the kinds of problems involved in mining bitcoin.

If you never mine any bitcoin, are your tools being used to mine bitcoin?

The bitcoin market is big enough, and there are enough places where power is free and the hardware is already paid for, that I am quite certain there's at least a few people still mining bitcoin today on GPUs.

Then there's the educational market, where it's less important that people mine bitcoin to earn money, and more about learning how bitcoin works by actually running the code.

So, no -- I don't think that's actually an open question at all. I don't have any proof, but then neither do you.

And all I'd have to do to create my proof would be to spin up my own machine with a GPU and use it to mine bitcoin for a while. Maybe on a free AWS account or something. Not that I'd actually do that, but it would be possible.

Name one piece of software that mines BTC on GPUs that has been updated in the last few years.

> and more about learning how bitcoin works by actually running the code.

You can find an old CPU miner for that. There's a few that have been written clearly with demonstration/learning in mind. If you don't expect any gains, why even go with a GPU?

Maybe because you want to study how Bitcoin works on a GPU?

Then they'll play with https://github.com/FourDizzle/js-gpu-bitcoin-miner/blob/mast... or similar on some custom chain or simulated environment. They wouldn't be doing any actual mining.

Maybe. Maybe not. But you can't force them to do so, even if that is the only way to save your theory.

Maybe they want to play with real software on real hardware. And maybe they've got access to free power to do it, or they already pay enough for power that the cost of a single GPU is not material to them.

> The bitcoin market is big enough, and there are enough places where power is free and the hardware is already paid for, that I am quite certain there's at least a few people still mining bitcoin today on GPUs.

> So, no -- I don't think that's actually an open question at all.

But you didn't even address my question! It doesn't matter whether someone is assigning their GPU to the job of mining bitcoin. Has any bitcoin actually been mined by a GPU? That would be the difference between "using a GPU to mine bitcoin" and "using a GPU to LARP mining bitcoin".

Can you prove that there is any Bitcoin throughout history that wasn't mined on GPU?

If that's what people are doing, then they have their reasons. And unless you can prove that they don't exist, then you can't rightly claim they don't exist.

That makes zero sense. If someone wants to play around with the concept of mining with a GPU, then mine ETH. Not only is there recent software that supports ETH mining on GPUs, but also the fact that it would take a long time to mine a single share on a BTC pool at this point.

Software for mining BTC on GPUs hasn't been updated in many many years. It would be a struggle to even get something working.

Maybe. Maybe not. Bitcoin is a better known name, and an easier algorithm to compare the differences between running on CPU versus GPU, and so that might be a very interesting thing to study.

I can definitely see a CS professor assigning that kind of work to their class for a while.

Then maybe a couple of weeks later, they do ETH instead. Or maybe Dogecoin, or whatever else the professor wants people to study.

For a hobbyist, maybe they really are married to Bitcoin on GPU, and damn the torpedoes.

The point is that there are billions of people in the world, and presumably at least millions of people who might be interested in mining Bitcoin. You can't account for every single person who might want to do that, nor can I.

My advantage is that only one such person needs to exist to support my thesis, whereas your position requires that there are no such people anywhere on the planet.

Like I said, these aren't going to be large bulk operations, I totally grant you that point. But I'm confident that there are at least a few examples out there.

Are you sure you really understand mining and the different algorithms used by the different cryptos? You don’t learn sha256 by putting a GPU rig that was only a transition between original CPU mining in the beginning and current ASICS mining.

> My advantage is that only one such person needs to exist to support my thesis, whereas your position requires that there are no such people anywhere on the planet.

Still nobody sane or competent would do it.

There are still edge cases where it could be done by people who are "sane". They're just interested in other things than you are. And they may have different cost structures than you do. But they're not "insane", even though they are making very little profit.

Brad, it would be literally zero profit to mine BTC with a GPU at this point. Even with pooled mining, the difficulty is so high that the odds of a GPU finding a nonce, even at a pool level of difficulty, is nearly impossible.

As I said before, there isn't even software available to mine BTC on a GPU (you never answered my challenge about naming it). It would be a lot of work to even get it running. So yes, it would be insane to even try at this point... just like doing it by hand on paper.

Fun thought experiment you're having here, but realistic, it is not. Definitely encourage people to try mining ETH with a GPU though... that opportunity is ending soon and it is quite interesting. Some GPUs have about a billion knobs in them to twist for tuning efficiency. They have silicon lottery effects on them as well, each chip is a snowflake.

I have so many GPUs that I've even written software to collect where each one was located on the wafer and done analysis of how that affected its performance. Spoiler, the ones on the edge don't perform as well.

In Norway until recently the electricity was cheap. So many apartments and even houses in suburbs use resistive heating as installing heat pumps was not profitable. But then in winter one can just mine crypto coins on GPU for free to heat the home. Granted the wast majority of miners work on alt coins but some do try their luck with bitcoins.

Stealing from society and burning fossil fuels in order to profit yourself.

It's disgusting.

How are they stealing from soceity?

This is achtually the nichest of niches where mining bitcoin is not stealing from Society: the existing alternative resisting heating, would produce as much heat per watt as mining bitcoin would.

Sure, there's the question of using the hardware for this purpose, and possibly this dampens the transition to heat pumps, but it's the 'least immoral' mining situation I've heard of so far.

Norwegian power production is to 90%+ hydro power. Norway is under the NO column in this table.

https://www.statnett.no/en/for-stakeholders-in-the-power-ind...

It's not in Norway that they are burning fuel to produce electricity…

The selling also has its own "difficulty adjustment". The miners who managed to sell their extra hardware early at the inflated prices made a good decision. Now that you can buy GPUs at below msrp again, there won't be a many exits.

The difficulty adjustment is deliberately slow (once every 2016 blocks) to prevent volatility in block hardness. That the hash rate is not dropping off a cliff over the last 6 months is testament to that.

https://www.blockchain.com/charts/difficulty

> to prevent volatility in block hardness

Kind of, it's to prevent an attack where nodes could be isolated and then duped to believe that the network difficulty is substantially lower than it should be. If we adjusted difficulty with every block (some altcoins do this), we could simply disrupt your connection to the network, wait until you would accept a block with a very low difficulty because 8 hours have passed with no blocks, and then mine that.

ETH adjusts every block. Your attack scenario is inaccurate. There is no single point of disruption to be had.

The block difficulty is calculated from the time between previous blocks. The time since you last received a block is not taken into account. Because not everyone clock is exactly the same, you could end up with different difficulties.

No, some altcoins actually do adjust like this to prevent their networks being "stuck" at a high difficulty.

You'd only get the isolated nodes to accept that block though, most nodes would reject it. So I'm not sure what you'd gain.

Many miners are currently finding themselves in an interesting situation.

Money was cheap. So if a miner thought that their cryptocurrency was going "to the moon", it was better to borrow money at low interest rates and stash the CC away for later.

Now that interest rates are increasing and the CC market is crashing, they cannot offload their CC quickly to cover their (future) costs and payments on their loans.

Is that true? Mining business has mostly fiat-based balance, and one doesn’t simply mix currencies in a business without hedging. I mean it is plausible that people could leave profits in crypto, but that’s just an intersection of a businessman and a crazy tothemooner.

It's a market. As bitcoin goes down, amount supplied of mining/hashing goes down. Difficulty adjusts downward which shifts the supply curve a little, but nowhere near enough to make it worth it to all the miners who mined before.

A few percent change in difficulty doesn't make a 50% drop in bitcoin all better.

Reply on news.ycombinator.com