Resource Center

Back to blog homepage

Comparing Consensus: PoW or PoS?


Batchable Computing, Bitcoin, Blockchain, Crypto Mining

By Dip Patel, CTO

Can we come to a consensus on consensus mechanisms?

When it’s about which approach is better for blockchain, apparently not. Get a group of crypto enthusiasts together, and one of the hottest topics of conversation that arises is the merits of Proof of Work (PoW) versus Proof of Stake (PoS). And there’s a reason for that: Once blockchain has been identified as the solution for a particular problem, the next big decision may very well be what flavor of consensus it should be — PoW of PoS?

A quick look back reveals how and why this lively debate has unfolded. The origins of PoW go back to 1999 when researchers strove to solve the problem of malicious (Byzantine) or malfunctioning nodes on a distributed network with the Practical Byzantine Fault Tolerance (PBFT) algorithm. Taking that one step further, PBFT was created to help achieve consensus or a process of agreement between distrusting nodes on a final state of data. Go from there, and you have consensus mechanisms, a set of steps that all — or at least most — nodes take to reach an agreement on a proposed state or value.

PBFT was an interesting concept that didn’t actually find a practical home until Satoshi Nakamoto launched Bitcoin in 2009, with PoW at its core. As its name implies, PoW has hard labor — of the digital kind — at its core, relying on proof that sufficient computational resources have been spent by the person validating transactions (that person is known as the “miner” in the case of bitcoin). Only then will the system propose a value that the entire distributed network will accept.

PoW worked wonders, powering Bitcoin (and therefore cryptocurrency and blockchain) to the forefront of global consciousness. But with significant energy consumption required by Bitcoin miners to come to a consensus and validate a transaction, alternative consensus mechanisms were sought.

The leading alternative to PoW emerged in 2012 with the proposal of PoS, which shifts the focus from energy-dependent miners to “staking,” where an algorithm selects nodes based on the number of coins (i.e. their “stake”) an individual has. In other words, this user has enough invested in the system that it shouldn’t be worth it for them to commit an act of malice — they have more inherent trust, in the form of coins staked in their wallet. As a result, they get a respective increase in opportunities for adding the block to the chain, thus reaping the network reward.

While the well-known blockchain Ethereum is currently PoW, its plan is to fork over to PoS in the near future. Top coins in the PoS world include Binance Coin, Stellar, Dash, NEO, and Cosmos.

Sizing Things Up: Decentralized?

Look to the market for a clear winner, and you’ll see that the jury is still out: The resource CryptoSlate lists 524 PoW coins, with 415 coins in the PoS camp — clearly, developers are putting plenty of time and brainpower into both approaches.

So how do you compare the two? The way I see it, the differentiators are driven by energy use, efficiency and security. While PoS is succeeding in reducing energy use/computational power and increasing efficiency by, in theory, providing a faster path to consensus and transactions, there’s a major tradeoff that comes with that: With PoS you’re effectively giving up decentralization for the sake of these efficiency gains.

Given that a major impetus for introducing cryptocurrency was to create an alternative to today’s highly centralized banking systems, PoS isn’t making the intended progress — in fact, it looks exactly the same as traditional banking. PoS works well for companies that want to deploy private blockchains, but for a public ledger reducing the number of validators to those who have stake in the chain, and therefore more voting power than those with a smaller stake, means you’ve gone back to increasing centralization.

Another way to think about it is our voting system in the USA: popular vote vs. electoral college. Back in the 1700s, an electoral college made sense for a young America, since a truly decentralized approach (popular vote) was overly impractical due to an inability to comprehensively share information with constituents and educate them on the issues, as well as accurately collect everyone’s votes. The introduction of the electoral college centralized the vote to an extent, as a critical number of votes in a state (excluding Nebraska and Maine, which have their own systems) lead to a winner-take-all casting of a state’s electoral votes for that Presidential and Vice Presidential candidate.

A parallel situation has developed in blockchain, where we have a pure, decentralized popular vote in the form of PoW and now are trying to make it more efficient by making special nodes that are given a larger voice via PoS. Unfortunately, choosing to achieve consensus via PoS leads to the aforementioned effective re-creation of the existing banking system — only now blockchain is being used as the security protocol instead of the banks’ backends. As PoS progresses, it will look more and more like traditional banking since “special nodes” will have most of the rights. For example, in Libra’s case, Facebook and the rest of the conglomerate represent “Special Nodes” — not necessarily people I trust to have my best interests in mind!

Evolution at Work

The most frequent point of contention between PoW and PoS is the energy consumption component. Undeniably, Bitcoin’s PoW mining process is energy intensive. It’s a drawback that Soluna is converting into a strength: Building high-performance off-the-grid data centers that are co-located with, and powered by 100% renewable energy power plants. In this way, the high power consumption of these blockchains actually becomes a profit engine, allowing green energy plants — which often have a track record of being unprofitable — to be constructed and become profitable through our unique approach called vertical integration.

“The relative energy efficiency of PoS is a valid benefit, but as noted above, it comes with a cost to security and centralization that can be self-defeating in blockchain.”

Often overlooked in the energy use comparison is the fact that virtually all groundbreaking technologies consumed a lot of power early on, and then got to a much greener place:

  • Traditional incandescent light bulbs have given way to more efficient, longer-lasting LED lights.
  • Cathode ray TVs and computer monitors have evolved into environmentally superior LCD (and now LED) screens.
  • Automobile engines have come a long way, burning cleaner and cleaner with the eventual graduation to an all-electric fleet ever closer on the horizon.

We can count on PoW to follow this same positive trend. As the flagship application for distributed ledger technology, Bitcoin is undisputed in its ability so far to deliver on the best of what blockchain has to offer: decentralization, transparency and trust, immutability, high availability, extreme security, and cost savings. Not to mention the utter disruption of financial paradigms long thought to be untouchable.

PoS could get to that same happy place, someday — but how much energy will developers expend to make staking as practical as its counterpart in consensus? Making PoW as efficient as possible just may be the better bet.