In addition to our focus on fees reduction, we are also fully committed to achieving another objective:
To become the L2 with the highest TPS capacity

Our ultimate goal is to be able to handle thousands of TPS to increase the congestion threshold 🧑‍🔬

The next step towards this goal, Transaction Parallelization, is coming this quarter 👷‍♂️

This new mechanism will enable the Starknet sequencer to process unrelated transactions in parallel rather than sequentially:
🐺 Sequential transaction management processes transactions one after the other, regardless of their (in)dependence.
🐺 Parallel transaction management processes unrelated transactions (such as sending ETH from A to B, minting an NFT collection by C, and swapping USDC for STRK by D) at the same time, optimizing TPS capacity

This is a huge step for Starknet, as we will be the first L2 to implement such a mechanism, and it's happening very soon (Q2)! ⚡️

Last week we unveiled a new standard we are working on: ZKThreads 🧑‍🔬

This standard, being developed in collaboration with @cartridge_gg, introduces a novel framework for execution sharding that provides scalability and customizability without fragmentation. It also allows for direct fallback onto the base layer, ensuring continuous operation at every step.

In simple terms, ZKThreads will offer developers improved:
✅ Security: Developers can execute directly on Starknet when necessary, ensuring continuous operation at every step and a leaner design for dApps.
✅ Composability: Developers are provided with a standard, composable toolkit to create a wide range of provable applications, including native bridging, common DA layers, and security mechanisms.
✅ Interoperability: All applications using the ZKThread standard will seamlessly interact with each other, reducing fragmentation and opening up numerous new use cases.

For users, ZKThreads will bring:
✅ Enhanced UX: Thanks to the interoperability provided by ZKThreads, users can switch between apps built as ZKThreads without changing networks or other infrastructure.
✅ Lower transaction fees: Similar to L3s, ZKThreads allows for further scaling on Starknet by adding an additional layer of proof compression before settlement, resulting in even lower costs than on the public Starknet L2.
✅ No fees on the thread: Apps operating as a ZKThread update the base layer in bulk at regular intervals. This means users won't have to pay fees on their individual transactions within the thread. Fees only apply when settling on the base layer.

Yes fam, this is huge!

Learn more by reading the following article and stay tuned for even more!

A few weeks ago, we announced the definitive targets of our 2024 roadmap:
🐺 Becoming the L2 with the highest TPS capacity
🐺 Average transaction fee consistently below $0.01

Now, let's discuss how we plan to achieve these targets! 👇

1⃣ Becoming the L2 with the highest TPS capacity thanks to Transaction Parallelization and Cairo Native

In order to fully understand both upgrades, here is a quick reminder: the mission of the Starknet sequencer is to process transactions and create blocks. Hence, any optimization of the Sequencer increases the network throughput. This is exactly what Transaction Parallelization and Cairo Native offer:

🐺 Transaction Parallelization allows the sequencer to concurrently process several transactions.
🐺 Cairo Native facilitates faster execution of Cairo instructions by circumventing an intermediate compilation step.

The principle behind Transaction Parallelization is very simple: rather than managing transactions sequentially (one by one), the sequencer will be able to manage transactions that are unrelated to each other at the same time. Crude example: an ETH/USDC swap on AVNU by user A with an NFT mint by B on Pyramid.

This will allow the Starknet sequencer to be able to handle many more transactions at the same time, reducing latency and increasing throughput 🕺

On its side, Cairo Native is a compiler with the goal to convert Cairo's intermediate representation (called Sierra) into Multi Level Intermediate Representation (called MLIR). In simple terms: Cairo Native makes it easier to understand Cairo instructions.

By integrating Cairo Native into the sequencer, the sequencer will be able to skip decoding & running cairo-vm instructions, and run native assembly directly, thus allowing it to execute transactions and create blocks more quickly!

Here is the superb analogy of @espejelomar to easily understand the implications of Cairo Native:
https://t.co/WXgNL4DzYh

2⃣ Reduction of DA costs through Volition

We have several optimizations planned at the stack level throughout 2024 to decrease fees and costs.

As you know if you're paying attention (🤓), Starknet’s main operating cost arises from the data we store on Ethereum, which is related to transactions made on Starknet. The EIP-4844 upgrade, along with various optimizations around it, resulted in a massive 95% cost reduction (!) on Starknet. This great cost reduction came with a corresponding reduction in transaction fees.

It’s huge, ngl, but we believe we can further reduce transaction fees by another order of magnitude this year. The main central feature in this effort will be Volition.

Volition will give developers the option to choose where the transaction data for their dApps are stored:
🐺 On Ethereum, which offers maximum security but for a higher fee (Rollup mode)
🐺 Directly on Starknet, which remains highly secure, albeit slightly less than Ethereum, while being significantly cheaper (Validium mode)

This will allow developers to optimize their contracts by choosing the data availability mode at the individual storage variable level. In this way, the trade-offs regarding gas fees associated with a dApp, and its level of security, will be determined by developers and users of each dApp themselves.

Note that Volition's timeline is still TBD.

In conclusion, the main takeaway is that Starknet aims to be the cheapest L2 to use on the market and the fastest by a significant margin. And our plan to get there is crystal clear:
🐺 Transaction Parallelization and Cairo Native to substantially increase our TPS capacity.
🐺 Volition to further reduce TX fees by a new order of magnitude.

Exciting time ahead 🧑‍🔬