Holochain removes the need for global consensus, and with it the expenditure of massive amounts of electricity (proof-of-work or ‘mining’) or accumulation of capital (proof-of-stake) used to ensure data integrity across millions of nodes about data that may or may not be relevant to them. This has three effects:
- Overall storage requirements are reduced, preventing the need to manufacture storage devices.
- Consensus protocols such as proof-of-work are unnecessary, preventing the manufacture and early disposal of yet more specialized hardware and requiring much less electricity.
- Network chatter is reduced, reducing the need for more infrastructure.
In fact, Holochain’s unique ‘immune system’ approach to data integrity means that the hardware with which end-users access the distributed network can also be the hardware that validates and replicates data. This is a contrast to popular blockchains, which typically need a large hardware investment in order to become a full participant.
While Holochain certainly compares favorably against most blockchain tech, which is already laudable, how well does it fare against the status quo of internet technology? I.e. HTTP and centralised servers.
Partner project RedGrid is very directly using Holochain to achieve greater energy efficiency, so that’s already very encouraging.
But I’m also under the impression that Holochain could be the more energy-efficient alternative for more general internet use as well. One aspect of it is the shorter distance bits need to travel because of p2p. But theoretically we’d also need fewer servers, because we’d be making better use of the ones we already have at home. (Though, unfortunately, when we have an abundance of something we tend to respond by using it in excess).
Should anyone like to venture some napkin math on possible outcomes I’d be all ears
Example scenario: Let’s say all chat services in the world right now were built on top of Holochain. Would there be a noticeable difference in energy consumption?
This is a tricky question, not easy to answer. Factors involved:
- there are efficiencies of scale in data centres, and they’re incentivised to make things even more efficient by balancing load across machines for optimal usage
- however, data centres spend 50% of their energy on cooling alone
- data centres are obsessed with reducing power consumption in server and networking hardware
- however, this means hardware, which represents increased energy consumption in some factory in Shenzhen, along with e-waste
- running a node in your basement, the heat is minimal and gets dissipated in a large envelope
- additionally, it might even reduce your heating bill by a few cents a day
- however, you’re using energy that you wouldn’t have otherwise, and with less efficient compute and networking hardware
- and that little P2P node is another piece of hardware that wouldn’t have been manufactured otherwise
- a P2P system can reduce the strain on global networking infrastructure if it’s only used by people in a single office building or town
So you can see that there are lots of pros and cons both ways.