Scaling Golem: research on scalability alternatives update

Hello everyone! It’s been a while since we addressed this issue. As we mentioned before in our AMAs, this was mostly due to the fact our team was focussing on more pressing stuff, and waiting a short while until scaling alternatives evolved.

As you might also know, we are currently preparing a migration to ERC-20, and hired ETHWorks, one of the most recognized dev shops in the market, to help us out. They have been working on great projects, such as Universal Logins, which is also being considered to improve UX after the migration. We will update more on all of this later, as the work is ongoing and we are happy to say it’s finally starting to take shape.

One of the Migration requirements was that our new ERC-20 should be compatible with Layer 2 scaling solutions. To this end, we have been reaching out to various teams working on L2 and commissioned ETHWorks to do research sprints in order to find out which alternatives could fit Golem best.

So far, there are a few promising projects that we are actively researching, so we wanted to share them with you. We want our community to be informed every step of the way, so this is a good opportunity to show our commitment to scaling and improving Golem is stronger than ever.


ZKSync is the ZK Rollup (layer 2) architecture by Matter Labs. ZK rollup features off-chain computation + on-chain data + validity proofs (ZK-SNARKs / ZK-STARKs). This approach helps solve the data availability problem inherent to Plasma. Even though it’s not yet ready to implement for general-purpose smart contracts, it can already be used for token transfers and specialized applications.


StarkDEX is a scalability engine for DEXes (non-custodial exchanges) that will allow them to handle over 10,000 trades per block (it is important to note, from my personal understanding that this is only if EVERY transaction in an Ethereum block uses StarkDEX). Has similar Data Availability challenges as Plasma. Starkware, with regards to their project, mentioned: “Data availability [is] being guaranteed by a trusted committee of participants (an example of potential participants: 0x, StarkWare, DeFi participants, etc.). Down the road, we (Starkware) plan to replace it with a blockchain-based solution.”

While we have been researching many other alternatives — for instance, Raiden and Loopring — these two, ZK-Sync and Starkware, are the ones that we’re focusing on right now. As we refine our research and the ERC-20 migration advances, we will be updating you more on the progress of these. If you have any other ideas, feel free to drop them on Reddit!

Originally published at on March 24, 2020.