Bitcoin layer-2 scaling networks—particularly “rollups”—have been the talk of the town in crypto developer circles as a new means to keep Bitcoin payments cheap, fast, and decentralized. But despite their vaunted promise, data and analysis from Galaxy Research suggest the vast majority of Bitcoin rollups will be unsustainable.
“Rollups on Bitcoin that post data to the base layer will face a significant problem: the cost to post data,” wrote Galaxy analyst Gabe Parker in a report published Friday. “Rollups on Bitcoin will need to generate substantial revenue from transaction fees on their own networks, driven by sizable numbers of users paying to transact on layer-2.”
Rollups are off-chain execution environments where transactions are “rolled up” and later settled in batches on a more decentralized, more secure blockchain. Ethereum has used rollups like Optimism and Arbitrum to scale for a long time, but developers only recently unlocked the technology to build rollups on Bitcoin.
Bitcoin rollups will use the Bitcoin blockchain as a “data availability layer.” That means they’ll post enough data to the blockchain so that anybody running an ordinary Bitcoin node can reconstruct the most recent state of the rollup network at any time.
Bitcoin blocks only have 4MB of storage capacity, however, and posting data to Bitcoin requires a lot of data. Specifically, each data posting will require zero-knowledge proof outputs and state differences—the latter of which scale in size alongside transaction volume and type.
bitcoin is an inflexible system to which all must bend the knee: critics, governments (!), and coders
zk-rollups are not exempt.
blockspace is capped (4mb per block) to promote the decentralization and robust fee markets. bitcoin was never made for data availability at scale pic.twitter.com/brGXnRMNpU
— Alex Thorn (@intangiblecoins) August 2, 2024
“Each individual data posting transaction can consume up to 400KB (0.4MB) of block space, effectively occupying 10% of an entire block,” wrote Galaxy. As there will be multiple rollups, which are expected to post their data every 6 to 8 blocks, rollups could quickly drive base-layer fees to new heights and price out smaller transactions.
Given the competition for block space, only the rollups that generate the most fee revenue to pay their way into blocks will be able to stay afloat.
Galaxy estimated that in a low-fee environment, when ordinary transactions cost 10 sat/VB (10 satoshis per vByte, a unit of block space data), rollups will accrue $460,000 in monthly expenses to afford Bitcoin’s security. High-fee environments of 50 sat/VB—as seen when Ordinals or tokens minted via the Runes or BRC-20 standards are seeing high activity—can drive monthly costs to $2.3 million.
Granted, fee estimates among different rollup systems are still highly variable—especially since no real rollups have launched on Bitcoin yet. BitcoinOS, which claims to have verified the first zero-knowledge proof on Bitcoin, told Decrypt in April that it expected to initially achieve 10x scale for transactions.
The degree to which these costs translate to rollup users also depends on how much activity the rollup has gathered: The greater the volume, the smaller the cost for individual transactions.
“It’s worth noting that estimates for zk-proof and state difference sizes are continually evolving as teams research and optimize data compression mechanisms,” Galaxy said.
One burgeoning Bitcoin rollup system is called “Build on Bitcoin” (BOB), a hybrid rollup meant to connect to both Ethereum and Bitcoin. Right now, BOB is merely an Ethereum layer-2 that offers fast, virtually free transactions, but will later upgrade so that it plugs directly into Bitcoin as well.
Alexei Zamayatin, co-founder of BOB, believes Bitcoin rollups can be just as cheap as Ethereum rollups—but that they shouldn’t actually use Bitcoin’s main chain for data availability at all.
Report on L1 data costs on Bitcoin from the perspective of L2s.
My 2 cents (imho this is missing in the report).
– No-one will post data on the L1. Use @CelestiaOrg or Bitcoin-native solutions like @babylonlabs_io @nubit_org , or a merge-mined DA sidechain.
– Bitcoin L2s will… https://t.co/IU8pZ9us4s— Alexei Zamyatin | Hiring (@alexeiZamyatin) August 2, 2024
Instead, he recommends using Celestia or a merge-mined Bitcoin sidechain for this purpose—an option that’s cheaper but loses Bitcoin’s full decentralization and security as a tradeoff.
“No one will use Bitcoin L2s if they are 100x more expensive than Ethereum L2s, just because ‘it is on Bitcoin,’” wrote Zamayatin on Twitter on Friday in response to the Galaxy report. “Good news: They won’t be more expensive.”
Edited by Ryan Ozawa
