Planning of the Gas Model with Separate Storage and Execution Fees for Aptos

According to the latest news, the Aptos team is actively promoting the Gas fee improvement plan. Through in-depth communication and feedback collection with the community, they have developed a phased improvement plan that is expected to be gradually implemented in the coming months.

This improvement plan mainly includes three phases:

1. Early January 2023: Reduced the Gas price of dynamic NFTs to one-tenth of the original.

2. Q1 2023: Build gas-efficient data structures and provide comprehensive support from guides to indexing.

3. End of the first quarter or early second quarter of 2023: Design a new Gas model that separates storage and execution costs to achieve demand-driven execution Gas costs.

Background Analysis

The initial goal of the Aptos team was to reduce Gas fees by 10 to 100 times. To achieve this, they engaged in in-depth discussions with multiple ecosystem builders and analyzed a large amount of on-chain transaction data. The research found that the token transfer fee on Aptos is approximately 0.00055 APT, while the fees for many DeFi operations range between 0.003 and 0.005 APT. Among these, execution costs account for more than 50%, while project creation costs account for more than 20%.

However, there is a key issue with the existing Gas policy: Aptos bundles the execution and storage Gas costs together, despite their fundamentally different roles. For execution, Gas limits the maximum execution time of transactions; while for storage, Gas determines the intelligent use of scarce resources. This bundling leads to a situation where reducing costs on one side affects the other, thereby limiting the overall optimization space.

Short-term Plan: Reduce Dynamic NFT Gas Costs

Although it is not possible to immediately reduce the overall Gas fees by 10 to 100 times, the team has identified a domain that can quickly make an impact: dynamic NFTs. By redesigning the implementation of SimpleMap, Gas prices can be reduced by more than 50%. This update is currently being prepared for the mainnet and will be released as the first Aptos Improvement Proposal (AIP).

Medium-term plan: Efficient Gas Data Structure

Many ecosystem teams have realized the nuances of storing Gas and have developed their own efficient Gas data structures. The Aptos team is also exploring this area, considering factors such as data size and application scenarios.

They are developing SmartVector to optimize data aggregation while exploring SmartTree, which has efficient lookup and incremental operation times. In addition, the team plans to restore support for BucketTable for more efficient large-scale map-like storage.

These improvements will be gradually rolled out through AIP, documentation, and index support. It is worth noting that much of the code for these improvements is already available, allowing developers to benefit without waiting for an official release.

Long-term Plan: Demand-Driven Gas Cost Model

The current Gas framework combines execution fees and storage fees, leading to an imbalance in Gas prices. In addition, the existing storage pricing scheme fails to reflect the data lifecycle and lacks incentives for deleting data.

To address these issues, the Aptos team plans to: in the coming months.

1. Separate storage and execution Gas fees, and reduce execution costs.
2. Establish a framework that supports the refund of storage fees.

Although the storage fee refund plan is still in the early conceptual stage, some core principles have been established, such as refunding storage fees to the original account that paid for the data creation and allowing subsidies for data creation from a central account.

After these improvements are completed, it is expected that the Gas fees for NFT operations, oracle updates, and DeFi-related transactions will be significantly reduced.