The alarms are sounding: Cryptocurrency is an environmental disaster in the making, a burning tire fire raging at the precise moment we need to be making significant strides in the opposite direction. Oceans are rising, droughts are worsening, and hundred-year storms are blowing in every five years.
To remain viable, crypto must find a way to be part of the solution instead of the problem. The good news is that far greener options exist and are already proving their worth for the next generation of blockchain technologies like The Internet Computer.
Bitcoin’s Energy Problem
Bitcoin’s dramatic rise to a global currency powerhouse comes with a cost: As Bitcoin’s value has soared, the work occurring to mine, validate, and transact on the blockchain has scaled exponentially, making the process extraordinarily resource-intensive. The New York Times estimates that mining bitcoin currently uses 91 terawatt-hours of electricity per year — the equivalent of what the entire state of Washington uses in a year.
The root issue is Bitcoin’s “Proof of Work” model, which was never designed to operate at this current scale. How does proof of work … work? It all centers around the blockchain, a centralized ledger that tracks all transactions and is maintained by independent contributors to the system. “Miners” — computer servers that record to the central shared ledger — must solve a complex cryptographic math problem. Miners who guess the solution to the problem first are rewarded with bitcoin.
Ten years ago, mining involved a handful of computers running under a desk or in a dorm room. Today, as the rewards have gotten significantly more competitive and valuable, mining happens in giant server rooms around the globe requiring constant cooling. The designed redundancy of everyone competing to solve the same math problem creates security and a colossal vortex sucking in electricity. There has to be a better way.
Proof of Stake
Where Proof of Work is a competitive system, with each miner attempting to solve the crypto math first and win the lottery, “Proof of Stake” creates a collaborative system. “Miners” are replaced with “validators,” who, instead of doing redundant work, take turns validating each block and writing it back to the blockchain. Validators must hold a purse of coins to participate — the “stake” in Proof of Stake. Correctly validated blocks are rewarded with more coins; mistakes are fined. This creates a monetary carrot-and-stick system.
The stake each validator holds is made of collaboratively pooled coins, providing profit to the investors putting up the cash to fund the system. The result: an incentivized system of validators that work collaboratively instead of competitively, eliminating a considerable amount of the redundant work — and electrical usage — required by the older Proof of Work system. All of this is accomplished while maintaining the blockchain’s core self-validating system, ensuring a trustworthy blockchain.
So how much better is Proof of Stake for the environment?
To help answer this, I spoke to Alin Sinpalean, a software engineer with the DFINITY Foundation, developers of the Internet Computer (ICP). ICP is built on a modified version of Proof of Stake, providing its blockchain with the same security and light power consumption.
Sinpalean did some back-of-the-napkin math to help break down the issues at play, including cooling overhead for servers, the efficiency of each system as load goes up and down, as well as redundancies necessary to ensure that any one server isn’t a single point of failure.
Sinpalean built his calculations using “pessimistic” assumptions, rounding up electrical usage where possible and assuming huge transactions taking up one second per transaction, or 1 billion instructions. (He emphasized these are not measured final numbers.)
Update from Jesse 4/15/22 : This article originally contained a mistake regarding the math for ICP’s electrical usage. I apologize for the confusion, and am happy to provide an update with more clear math. Alin wrote in an update:
It was brought to my attention that people are quoting your article and pointing at just how inefficient the IC is. I guess the misunderstanding comes from the confusion between kW and kWh. (kW is equivalent to mpg; kWh is equivalent to gallons of fuel. One would measure fuel consumed to get from A to B in gallons, not mpg.)
The 1 kW I quoted was my estimate of the rate at which a server consumes electricity. It is equivalent to a miles-per-gallon number for fuel consumption. The IC doesn’t consume 13 kW per second (i.e. kW / s), it consumes 13 kilowatt-secondw (kW * s) to execute 10 transactions. (In the worst case scenario (of a transaction running the maximum allowed number of instructions. Which is not what we are seeing in practice.)
1 kilowatt-second is 1/3600 kilowatt-hours (kWh). So my worst-case estimate for the IC’s energy consumption (10 canisters per subnet, each running the longest-possible transaction; 13 replica subnets, each requiring 1 kW of power) is 13 kilowatt-seconds for 10 transactions, or 13 / 3600 / 10 = 0.000361 kWh per transaction. Worst-case scenario. Not 1.3 kWh. Similarly, Bitcoin’s power consumption is 707 kWh, not kW per transaction.
another DFINITY team member offered:
Another way to think about it is that the subnetwork processes 36,000 transactions per hour at a cost of 13kWh (13kW of power for one hour is 13kWh of energy) and 13 / 36000 = 0.00036111111. Bitcoin by comparison uses 2131.42kWh per transaction (1000s of servers computing hashes for several minutes). This is 5'902'393 times more. Not 543 times more, as the article claims.
For comparison, Bitcoin currently uses 707 KWh — that’s 543 times more electricity to operate than the Internet Computer. Ethereum, another Proof of Work competitor, is positively sipping on electricity compared to Bitcoin, needing 62.5 KW — 48 times more electricity than ICP.
Cardano, another Proof of Stake system, which prides itself on being super energy efficient, reports needing just .5 KWh. Solana also recently reported an impressively meager number, reiterating the proof of stake‘s efficiency. Both are less than ICP but include fewer services. ICP’s system provides web hosting and memory, compared to Cardano, and Solana-based apps would have to outsource to another system, such as AWS, further adding to their energy footprint.
All told, it’s an impressive report card for the Internet Computer’s carbon footprint, and most importantly, it can scale as the system expands, getting more efficient with time and growth. That built-in scalability is one of ICP’s strengths.
A Stake in Our Future
Just as national borders or traditional limits don’t bind crypto, the climate crisis spills over from one country to another, impacting us all. Cryptocurrency’s future depends on a green approach to energy usage — one that understands the effect of growing these systems on a global scale and embraces the challenges presented by climate change.
The implementation and success of Proof of Stake are vital for the future of all blockchains. Systems like the Internet Computer that embrace this green future will thrive in this environment. We are all staked and pooled in the future of our planet. Let’s make it a good one.