# Ethereum's scaling roadmap: why current plans are too unambitious

Source: https://x.com/dankrad/status/1881405119373234238
Kiwi News: https://news.kiwistand.com/stories/Ethereums-scaling-roadmap-why-current-plans-are-too-unambitious?index=0x678eb3077e63e0274a1b5d14e5461565f955ad0645e8c5a29f5ee457d8d61063a61f8cd0
Submitted by tim on 2025-01-20 UTC, 2 upvotes, 0 comments

## Summary

Ethereum's core development process is described as broken, and the author argues the current scaling plans are too unambitious. The post says builders need long-term capacity commitments and five-year plans, and it calls for agreed timelines in the core dev process. It argues that L1 scaling by 100 to 1000 times is possible within a few years, citing Vitalik's 2021 roadmap post and the arrival of DAS and zkEVMs. It also proposes reaching one megabyte per second for data availability in 1 to 2 years rather than 5, while noting that other data availability protocols are building for GB/s.

## Article

Some points on this:

1. The current process of upgrading Ethereum IS BROKEN.  Builders need long-term commitments to capacity, otherwise they cannot build on Ethereum. They need to know what the plan is in 5 years.
So I hope that we can finally agree to timelines in the core dev process.

2. We nonsensically claim that L1 cannot be scaled and L2s are the only way.
This is a complete ivory tower perspective -- Vitalik already posted this 3 years ago, and with DAS and zkEVMs the ingredients are now ready: https://vitalik.eth.limo/general/2021/12/06/endgame.html
There are limits to this, yes, but scaling it 100-1000x in a few years is completely possible.
This achieves many things that based/native/etc. rollups don't.
I think actually doing this and stopping to find excuses for not doing it is essential for Ethereum winning.

3. Not everything can be on Ethereum L1, whether the actual limit is 1k or 100k tps or maybe even 1M tps. For this it is still great to add lots of data availability and create a great Ethereum security zones for rollups, who are Ethereum allies and can use Ether and all other Ethereum assets trustlessly.
But we need to do it at a scale that actually matters. We need to get to our current target of one Megabyte per second in 1-2 years rather than 5, and then KEEP GOING. I am currently doing research on maximally decentralizing the networking DAS/Data Availability Sampling, but until that's done, adding Data Availability Provider nodes with higher requirements is a good solution that has the same security assumptions as optimistic rollups (one honest node).
Other Data Availability protocols are currently building for GB/s! Ethereum does not have to match this, but being at 1/1000ths is not good enough.
Some projects are openly stating already that they will never use Ethereum DA (and openly advertising with Vitalik's support! The irony). Being ok with that and still claiming that they are somehow "Ethereum aligned" is thoroughly un-ambitious.