SHA256: Larger Bitcoin blocks could make mining more expensive

It seems like the large block discussion has been all but pacified, but there’s an issue that didn’t receive as much attention as it should have: the mining cost for larger blocks.

SHA256 Computing Cost

From quick (and admittedly not thorough) research we find that the SHA256 algorithm runs in approximately O(n) time. This means that the time needed to process data using this function grows roughly in a linear fashion. So, as you double the amount of input data, the time needed to process this data will also approximately double.

Note: The process of computing a block’s Merkle Tree only happens once during mining. So the bigger data processing cost, as described in this article, only happens once per block until a header is built. After that, miners will continuously hash the same header over and over. The larger block size does not incur in additional costs during this repetitive phase.

Energy Cost

There’s another linear relation involved in Bitcoin mining: energy price. The electricity bill also approximately doubles when you double the power consumption. There are additional considerations, but let’s isolate these two pieces of information for a moment.

If Bitcoin blocks were made 8X larger, as was proposed by some, the SHA256 Merkle Tree calculation would consume 8x more energy per block.

Miners only work on the block header, but every mining pool has their own idea of what a block should be. So, globally, until a block is accepted, an arbitrary number of miners would be assembling different larger blocks for the first time.

August 2020 Update

It’s been exactly 2 years now since Bitcoin SV was announced. Ever since the Genesis hard fork from February 2020, BSV has no hard limit for block size.

In all this time, Bitcoin SV blocks have naturally remained in the 2 to 4 MB size region, even though there’s no practical block size limitation in place.

Bitinfocharts shows us how Bitcoin SV block size has rarely been larger than 4MB:

Despite some monster blocks being mined here and there, we can see that there really wasn’t much demand for larger block sizes in BSV.

Conclusion

In short, making Bitcoin blocks 8X larger would:

1. Make electricity bills larger;
2. Require more processing per block;
3. It could increase the mining difficulty as well.

Larger blocks wouldn’t make a difference for individual miners who’re connected to pools. The pools, though, would incur in some extra cost while building the block header.

The idea behind larger Bitcoin blocks is to increase throughput. By increasing the number of TX’s in a block, you’d get more transactions every 10 minutes.

As we can see this isn’t necessarily a valid premise. When we last updated this article, a unlimited block size version of Bitcoin has been deployed and average block size lies between 2 and 4 MB, proving there isn’t much demand for larger blocks.

Links

Bitcoin SV Genesis Hard Fork – Feb 2020

Bitcoin SV data at Blockchair

[June 2019] Bitcoin SV (BSV) ‘Quasar’ Protocol Upgrade Continues Massive Blockchain Scaling, Lifting Default Block Cap to 2GB

Coin Dance BitcoinSV Block Details