Bitcoin SV has made the transition to (for all practical purposes) unlimited block size.
You can now store any amount of data on the BSV blockchain, which raises serious questions about scalability.
If every application decides to store data on one blockchain, will the system remain usable? Will setting up a full node continue to be viable for home users, as envisioned in the original Satoshi whitepaper?
The apps themselves may be decentralized, but the data storage from many different apps will use a single storage engine on a single blockchain. Is this scalable?
Will BSV mining still be profitable when the BSV blockchain grows?
Let’s take a look at some of these questions in a bit more detail.
What is the impact of large block sizes on Bitcoin SV mining?
To analyze mining performance we must look at two main factors: memory complexity and processing complexity of Bitcoin SV mining.
Mining: Memory Complexity
According to the top of the line S17+ Antminer manual, the Ryzen processor in each miner has 15 MB memory. Extremely low for today’s standards but still a lot of memory for SHA-256 hashing. Of course, the SHA-256 search does is not performed by the Ryzen processor but by the dozens of ASIC chips each of which has a tiny amount of memory reserved just to crank out hashes. The Ryzen’s memory is used to operate an operating system, a TCP/IP stack and to check for winning blocks and send the results out to the network.
This low memory consumption is the main factor behind SHA-256’s easy implementation in ASICs. (Several cryptocurrencies adopt hashing functions with a larger memory footprint to make ASIC construction more difficult.)
Therefore big blocks are not a big factor when considering memory consumption. SHA-256 is extremely lean and can process gigabytes using small chunks at a time.
Mining: Processing Complexity
SHA-256 computing time is roughly linear with input size or, in computer science terms, it’s approximately O(n).
Meaning that a 1 GB BSV block will take ~1000x more work to solve than a 1MB BTC block (1GB = 1000MB).
So, although mining machines will not require much more memory to process the large BSV blocks, they will require more processing power than Bitcoin Core whenever block size is larger than 2 MB.
Processing power is directly proportional to electricity costs. 1000x more computing cycles will cost 1000x more in electrical consumption. This is a very important factor, since Bitcoin mining profitability depends almost exclusively on electricity cost (the other main factors being equipment cost amortization and maintenance costs + downtime).
Bitcoin SV mining is a ASIC-intensive process which does not require much memory. Processing-intensive applications imply higher electrical costs.
When dapps being to store large amounts of data on the BSV blockchain, mining costs may rise linearly with the amount of data in each block.
This rising cost may not be viable for miners unless BSV coin value rises proportionately with data usage.
Everyone’s “tweches” are stored in the BSV blockchain.
Since there’s a cost to send a TX for mining, there’s also an incentive for users to publish : engagement with a post will generate fees which are shared between the network and content creators in completely decentralized fashion. Almost everything happens on-chain.
Sounds like a neat concept – to take control of your data back.
But there’s a problem.
Twitter produces 12 Terabytes of data every day!
If Twetch were to reach a small fraction of Twitter volume, the blockchain would become impossible to download for most BSV nodes.
For instance, if a cryptocurrency exchange were to host a blockchain that grew as fast as Twitter on Amazon EC2, their EBS (live hard drive storage) cost would grow U$ 1200 per day. At the end of just one month the live storage cost would be over U$ 36,000. And it’d grow by U$ 36k each month.
Lastly, there’s the question of immutability.
On Twitter you cannot edit tweets but you can delete them any time. On Twetch both things are impossible. Everything you post is immutable: cannot be edited and much less deleted. Years from now any mistakes you make online today will be available for auditing on the blockchain.
The immutability of data is precisely what makes blockchain ideal for financial solutions. Storing money on the blockchain, like Satoshi Nakamoto envisioned, is the killer application for an immutable database. But storing personal tweets, stories, photos and other content forever, without the possibility of rolling back, may not be a good idea.
We mentioned the Twetch use case because it’s a popular new dApp running on top of BSV, but this analysis applies to any similar system. Any system that stores raw data on the BSV blockchain will incur these same costs.
Development of large scale applications that store data on the blockchain is absolutely not scalable.
There’s illegal content stored in most blockchains. Bitcoin Core blockchain has lots of illegal links and binary data in it, BCH which forked from BTC does as well.
And it couldn’t be any different with BSV. According to the BBC, there are lots of illegal materials stored on BSV forever. You can’t edit the blockchain, you can’t remove content, you can’t clean it up – ever. The illicit data is stored in it forever.
We mentioned other blockchains which contain illegal material, so this isn’t specific to BSV. But here’s the catch: Bitcoin SV allows for nearly unlimited data to be stored per transaction.
While in BTC the attackers must limit their malware to a few bytes per transaction, in BSV they can publish a whole movie if the fees are high enough. This opens a whole new window of opportunity for wrongdoing.
And, again, once stored on the blockchain the content becomes perpetual. There is no way to edit a blockchain to delete the illegal material.
What happens, then, when a vandal decides to store violent, offensive and repulsive content on the blockchain? It will remain there forever.
This has deep legal implications. Take, for example, UK law where it’s a crime to possess any kind of child pornography material, regardless if you’ve ever opened it or not. Running a full node which contains the blockchain, which in turn may contain the illicit content, would be illegal in the UK!
There are several other variations of the UK example.
A blockchain that may or may not be legal according to the geographic location is definitely not scalable.
We’ve already touched on the subject of chain growth when we discussed the Twetch application.
There are other concerns when it comes to storage requirements.
For instance, what happens if a wealthy attacker decides to litter the blockchain with trash data? Let’s say a rival blockchain decides to spam BSV. How would dApp and full node performance be impacted if someone decided to pay the fees to push immense amounts of trash into the blockchain?
On Bitcoin Core this is not viable because blocks are limited to 2MB max (1MB nominal). When miners receive a bloated TX they first check the fee. If the fee is not satisfactory and the block is too large, the miners will ignore the block.
That’s not the case with BSV.
Bitcoin SV is designed for large blocks! Therefore BSV miners must accept large TX’s. Spammers are then free to submit immense TX’s into the network as long as they pay the fee.
Spam might become a big problem for the BSV blockchain and I would consider it not scalable.
On-chain Storage Cost
Let’s take this random BSV TX as an example.
In that TX, 14.798 kilobytes of data were transferred. The fee was 0.00003720 BSV, which at today’s prive of U$ 359, equals U$ 0.0117393 (1.1 cents).
One MB would be 67.58 times more expensive at U$ 0.7933.
So, to send gigabyte of data to the BSV blockchain, it would cost us U$ 793.30 at today’s prices when adoption is still very limited.
How can the BSV storage scale at these prices? It’s impossible.
On-chain storage cost on BSV is not scalable.
As you can see, having larger blocks could be a big bonus if only financial transactions were stored on chain.
From the moment you being to store BLOBs of all types on the blockchain, unlimited size blocks become a big problem.
We took a look at several factors including the cost to send data to the chain, to keep it live on chain, the possibility of spam and illegal content.
There are more factors to consider which we didn’t cover on this article, as the data storage possibilities are unlimited, especially on a publicly accessible medium such as a PoW blockchain.
I think BSV will not scale unless the type of data stored on chain is limited to financial TX’s.
Elephant scale image courtesy of Wikipedia