Monolithic vs Modular Blockchains: The Twin Suns Thesis
A New Hope for the blockchain scaling wars
While scrolling Twitter the past couple months, I’ve had the pleasure and displeasure of being forced to read up on developments in the blockchain scaling space. You’ve probably seen the propaganda and debates about modular and monolithic blockchains, it’s a narrative bloodbath out there.
Each camp thinks their tech scales better than the other. And it’s devolved into a narrative of monolithic chains VERSUS modular chains. A lot of “either or” rhetoric. It’s my chain vs your chain. It’s Solana vs Ethereum. My chain faster/cheaper. Your chain not decentralized. Hardware doesn’t scale. Rollups don’t scale. etc.
Despite all this noise, I’ve become pretty pilled on both approaches. But how is it possible to justify these seemingly opposite ideas? Don’t believe everything they say. Only a Sith deals in absolutes.
Let’s try something new (this is the tldr):
Monolithic chains and modular chains offer different tradeoffs for different use cases.
Both are useful. They will grow in parallel.
The real competition is monoliths vs monoliths, modular vs modular.
2 different battlegrounds. 2 different North Stars.
The Twin Suns thesis.
For some context, I got the idea for the name from Star Wars. The Twin Suns of Tatooine represent the dark side and light side, Sith and Jedi, among other things (some lore here for the Star Wars nerds). The scene also lines up smoothly for the header graphic, so bonus. Oh you think it’s corny? Could care less, you already gave me the click lmao.
What is Modular? What is Monolithic?
Monolithic blockchains
Monolithic chains vertically integrate the entire blockchain stack. So the chain does everything from processing transactions to ordering transactions to ensuring data availability. A blockchain that does all the blockchain things, pretty straightforward.
Side note, I don't think “monolithic” is the best way to describe it. Would say they’re more like “integrated” chains, but will stick with monolithic for now for recognizability.
Modular blockchains
Modular chains separate and specialize in different parts of the blockchain stack, such as rollups which focus on execution or Celestia which focuses on data availability. By separating the different components you can have different teams specializing and innovating in parallel. Think of each one as a lego block and you stack these specialized modular chains under the hood to build even more scalable chains.
It’s much more complex than that, so for more reading you should go straight to the modular propaganda source.
The Scaling Landscape
To make sense of the scaling landscape I came up with a rough framework. This is probably the part where I get get exposed as the mid curve larp that I am, is ok, will appreciate the feedback. Some disclaimers:
It’s focused on execution layers (the chains that process user transactions) here because they’re user-facing and it’s implied that the rest of the modular stack is working under the hood.
Yes it’s very oversimplified, generalized, missing a lot of nuance
ZK rollups are lumped with optimistic rollups here for simplicity, but are very different
The graphic is missing your favorite chain (highly encourage you to complain/reply guy me on twitter to boost my engagement)
So here’s my framework for thinking of the landscape right now:
Single-Threaded Monolithic
Monolithic chains that process 1 tx at a time. Most of these have moved to rollup/horizontal scaling roadmaps due to the limitations.
Ex: Ethereum, Polygon, Binance Chain, Avalanche
Parallel-Processing Monolithic:
Monolithic chains that process multiple tx’s at a time.
Ex: Solana, Monad, Aptos, Sui
Single-Threaded Modular
Modular chains that process 1 tx at a time. There’s more nuance here, not necessary for this post (see disclaimers)
Ex: Arbitrum, Optimism, ZkSync, Starknet
Parallel-Processing Modular
Modular chains that process multiple tx’s at a time.
Ex: Eclipse, Fuel
Landscape Takeaways:
It’s a scaling arms race, with single-threaded monolithic chains clearly falling behind parallel-processing monolithic chains like Solana on speed/cost/UX, which doesn’t bode well for mass adoption.
The introduction of modular chains/rollups provides an alternative scaling path, “horizontal scaling,” that hasn’t been fully explored. Optimistic rollups, zk rollups, appchains, are some examples. ZK rollups in particular still nascent but worth waiting for.
Most interesting to me are the upcoming parallel-processing modular rollups like Eclipse, which leverage the best parts of the modular stack and parallel-processing design to build chains potentially comparable on speed/cost to the existing parallel-processing monolithic chains.
Wait wtf, so is Nosleepjon pushing an anti-eth, anti-solana, pro-modular agenda? NO:
Monolithic and Modular Blockchains Will Coexist
I believe we’ll see both scaling strategies succeed, because they offer different tradeoffs for different use cases:
The Monolith Case: Simply Composable
Most users and developers will start with monolithic chains, it’s just easier, everything is already there.
Since it’s a vertically integrated system, you don’t need to worry about infra or anything except building your smart contract/app.
Monoliths are ideal general purpose blockchains. You have composability with the most onchain primitives, existing app ecosystem, existing community, largest pool of liquidity. Compare this to the fragmented state of horizontally scaling rollups.
It’s cheaper to build on a monolith. Most apps don’t need their own chain. Running and coordinating chain infra is expensive.
Gud reads:
https://multicoin.capital/2023/08/15/the-hidden-costs-of-modular-systems/
https://www.neelsomani.com/blog/a-year-around-the-sun-with-eclipse.php
The Modular Case: Sovereign Superapps
The term “modular” is actually a distraction tbh, the real benefit and theme is sovereignty/ownership.
You have more control/customizability, like with compliance, permissions, risk limits, forkability/upgradeability, direction of chain
Fat App/appchain thesis. A lot of value will accrue to a few apps rather than the chains they’re built on. We already see this on Ethereum with Uniswap, Lido, MakerDAO generating outsized fees. Or Friendtech generating multiples over Base. So some of these superapps will naturally want to capture more value with their own appchains. Especially applicable with consumer apps/games, less so DeFi due to breaking composability.
Easier and more customizable chain assembly
More monetization angles. For example you might be able to charge fee premiums on your appchain that you wouldn’t otherwise on a chain you don’t control.
Avoid noisy neighbor problems, guarantee uptime by running your own infra.
More evolvable, smartest teams constantly innovating on different parts of the stack. New modular tech can be swapped out over time
You think Coinbase doesn’t know what they’re doing? They control sequencer revenue and can drive their customers to their chain. Enterprise/web2 companies will love it. The sovereignty to control the range of permissions, monetization, build easily out-of-the-box.
Side note: It’s funny how we keep going in narrative circles with new tech but always end up back at the same Cosmos theses from years ago.
Gud reads:
https://zeeprime.capital/the-fappening
https://www.paradigm.xyz/2021/04/a-cosmos-thesis
Endgame?
Both cases make sense, and don’t contradict as much as the competing camps want you to think.
The Twin Suns endgame looks something like this:
Monolithic chains will end up competing with each other for general-purpose + composable DeFi use cases.
Modular chains will end up competing with each other for the sovereign superapp use cases.
1 dominant monolithic chain emerges, with a few smaller monolithic alternatives.
An interoperable ecosystem of modular/sovereign superapp-chains + smaller rollups/appchains.
1 dominant monolith because consumers naturally congregate around 1 category leader. Network effects help scale liquidity, user base, developer base, brand recognition, etc. Most efficient for pricing and UX to converge on one global general purpose monolith.
A few smaller alternatives, but most monoliths die out. Like the Uber/Lyft, Google/Bing type market share split.
A couple massive modular superapp-chains, Fat App thesis, more beneficial for these superapps to break off and start their own chains than pay rent to the monolith.
Network of longer-tail interoperable rollups/appchains (that rise and fall like apps on appstore?)
In this scenario, Rollups as a service providers become like VCs, exchanging limited engineering hours to service/bet on a portfolio of rollups/appchains. If one hits big, then fee/token share returns multiples for the provider.
Important not to underestimate the size of some of these future rollups/appchains. A consumer app like Friendtech with less than 100k users is generating more tx’s/fees than entire chains like Bitcoin, Arbitrum, BSC.
How do I Bet on This Shit?
My best guess:
The native token of the winning monolithic chain likely gains monetary status as the most recognized unit of exchange in the largest monolith.
The native token of the settlement layer for modular rollups/appchains likely gains monetary status as the most recognized unit of exchange in the modular ecosystem.
The dominant wallet and/or interoperability layer that abstracts away all these appchains into a seamless user experience. Likely have pricing power to charge a spread on all the apps they integrate because of the convenience. Maybe like credit card companies facilitating consumer/merchant interactions? Haven’t thought that far.
Final Thoughts
Make your own conclusions. Don’t let the talking heads pressure you into believing everything they say. Just because they're loud doesn’t mean they’re right (includes me).
I’m excited to see more experimentation with parallel processing rollups and zk rollups as potentially comparable performance-wise to parallel-processing monoliths. Not close yet though.
If we do have this dual ecosystem with modular/sovereign chains + monolithic chains, then we’ll need scalable interoperability (not that manual deployment + beg the bridge team political BS) to solve the fragmentation of liquidity and UX. It won’t just be about connecting modular rollups with each other, it’ll be about connecting the monolith to other rollups.
Mandatory DISCLOSED shill: This future is what Hyperlane (my bags) is built for. Gud read on Hyperlane here.
None of this is financial advice, you know the drill.
If you have any questions or complaints, I highly encourage you to reply under the tweet or @ me on Twitter. Cheers.