Bitcoin rival Chia 'destroyed' hard disc supply chains, says its boss
newscientist.com
newscientist.com
A typical plot is around 100GB and officially takes around 356GB (and a lot of SSD thrashing) to farm. Then that plot is copied off to long term storage (larger, slow hard drives via SATA, USB, etc) where it sits until a given plot of yours "wins". You can run one of these collections of hard drives off a Raspberry Pi.
Fun fact: In other crypto there are pools so that everyone contributing work gets a share of the profit. There are no pools in Chia. You're farming a digital lottery ticket that then sits there using up hard drive space and an always on PC hoping one off your tickets wins.
I researched it a bit so I'd understand the basics and the cost to earnings setup. It's so wasteful of hardware and space that I didn't want to set up a rig, though.
Can the thrashing of space be done on a RAMdisk, preventing SSD wear? Thereby only the 100GB would be moved onto long term storage, e.g. primary SSDs.
Is that worthwhile? Does RAM speed matter in this case?
The big problem is whether you can setup a 400GB Ramdisk?
I don't know if it still matters or it's too late now.
They measure lifespan in petabytes instead of terabytes. Or data writes per day (DWPD).
Now with this new currency, you could maybe even run cloud storage. Maybe we are accidentally building the futur alternatives to Amazon cloud services.
The chips that are mining bitcoin are not general purpose computers.
This is a part of why NVidia cracked down on gpu mining recently - when ASIC mining became more profitable than GPUs, the secondhand market got flooded with cheap high quality cards (thus ruining the primary sale market for new GPUs) and people flooded NVidia support with support queries for slow or unreliable GPUs.
When farming, disk speed is not critical, just size. Thus the idea of recycling millions of HDDs in a low power scenario. A few seeks here or there on your old, unused terabytes, but you need to have enough space to offer, otherwise your chances are zero. Today's farming is dominated by pools, some using non-official, closed source binaries of unknown origin. Official pooling is, supposedly, coming soon.
Plotting, which is preparing the files that are farmed, is, as of today, extremely IO intensive, but still cheaper, energy-wise, than mining. A single 100GB plot (you need ~20TB to have any chances of farming success) can take hours to plot. So people are resorting to extremely fast SSDs, or NVMe in RAID0 in PCIe 8x/16x cards to slash that time. And, in the process, wearing them out. Can't find the link now, but some were saying normal SSDs are ruined after 6 weeks of 24/7. Plotting keeps improving, although I'm not sure if it's got a lower performance bound by design.
From this initial organic buzz, speculative markets take over and the whole thing becomes a positive feedback loop.
Crypto is fundamentally a huge waste of advanced manufactured goods, electricity, and time, and enables tax avoidance and money laundering. The correct response from Governments is to simply ban the formal exchange of all of them in the same way that the USA bans online gambling and pyramid schemes.
So, alternative proofs of work still have an interesting basis for speculative valuation and acceptance. In the last crypto bear market, of the top 100 at peak, proof of work coins were the ones that had the smallest drawdowns.
finally someone calling it like it is
[1] https://cryptorials.io/beyond-hashcash-proof-work-theres-min...
I suspect nodes ask one another to perform computations on random subsets of very large amounts of data. Since the subsets requested would be random, nodes can't predict the requests and thus have to store all the data. The data would then be augmented with the consensus results of these computations, so that future computations would be able to reference them.
They are designed so that precomputing the bingo card is expensive but once you have them, checking to see if you have a bingo is cheap. It's thought that nobody will be able to compute a winning bingo cards on the fly ("grinding"), so it will be cheaper to store and reuse them once you have them.
(I don't know the details of the algorithms though.)
Then the transaction validation is done by whoever finds value in their storage (instead of doing a lot of computations online, they are pre-generated).
I assume there's a bunch of tree-like structures in the middle to make the search faster.
The trick is that calculating the response to a challenge takes a very long time and the server only waits for a very short time to get any valid responses. So it follows that the harvesters must compute the responses in advance and store them on disk, which serves as proof that you had that disk space allocated to chia harvesting.
you can't do that with miners, and GPU's are still fairly specific components. But plenty of companies and people have all sorts of needs for extra space.
I mean - if we're going to be scaling anything... hard drives are the thing. And if there are any storage optimizations people come up with, those will only benefit the storage industry overall; unlike ASIC miners, for example.
additionally, farming (serving plots) is done with regular drives, and those are very lightly read from and used once the plots are loaded.
https://www.tomshardware.com/news/chia-crypto-farming-can-de...
But I'm no expert so feel free to correct me if I'm wrong.
[I see from other comments that this is bad for SSDs but not so much for HDDs]
So 12 Exabytes? ¹
¹ or 11 Exbibytes
"Around 12 million terabytes of hard disc space is currently being used to mine Chia, having risen on an exponential curve since its launch in March. ... Hoffman says that if all hard discs sold in a year – which could hold data totalling roughly 1 zettabyte, or 1 billion terabytes – were used exclusively for mining Chia, it would still use less than 1 per cent of the energy currently used by bitcoin."
Quick, divide 12 million into 1 billion in your head! "Uhhh..." (It's 1.2%.)
Chia network is currently 12.04 EiB