Force Study | Before HyperDrive

In previous article, we have explained why HyperDrive upgrade on July 1 (Beijing time) was so important to Filecoin nework, which was due to that the reduction of the transaction fee in PreCommitSector and ProveCommitSector can improve storage efficiency.

This article mainly describes how to calculate the Gas fee after HyperDrive upgrade, taking the test result on Calibration network.

Calibration net: 0.2PiB storage power growth & 0.1 FIL total consumption per day on Filecoin network

In the Calibration network, Basefee is fixed at 100 nanoFIL, the current daily storage power growth of the whole network is about 0.2 PiB (starting to decrease on June 28 since most miners are expected to have finished testing), and the daily transaction fee is about 0.1FIL.

It can be assumed that the transaction fee will not followingly move up to hundreds of thousands with more than 100PiB of daily storage power growth, because during HyperDrive upgrade:

1. Gas fee will be extended from the original Gas fee of PreCommitSector to the Gas fee of PreCommitSector and that of PreCommitSectorBatch (whose is the cost generated by pre-submitting multiple sectors at one time).
2. Gas fee of ProveCommitSector will be extended to the Gas fee of ProveCommitSector, ProveCommitAggregate (which is the cost incurred by precommitting multiple sectors at once), and AggregateNetworkFee (batch fee for proof of replication).

As a result, the items of Gas fee consumption for the storage power growth after HyperDrivw are mulitiplied from 2 to 5 (PreCommitSector, PreCommitSectorBatch, ProveCommitSector, PtoveCommitAggregate and AggregateNetworkFee).

The following figure shows the distribution of Gas consumption:

Gas Fee Distribution on Calibration Network, from IPFSForce, 20210629

As can be seen from the above figure, the cost of transfer is huge. The total cost covered in PrecommitSector and ProvecommitSector is also reduced from 95% of the FIlecoin network to no more than 10% (calibration network).

So how much is the actual reduced Gas Fee in PrecommitSector and ProvecommitSector?

Gas reduction of calibration network is very obvious

Precommitsector decreased by about 50%

Total and Average GasUsed in PreCommitSector, from IPFSForce, 20210623

According to the data on June 23 from IPFSForce, the bulk pre-commit sector and the cost of the Precommit itself is the main reason of the decrease.

1. Pre-submitted Gasused is reduced by approximately 60%. According to the previous data, the average Gasused of the original Precommit Sector was about 50 million, and the current amount was about 20 million. Generally, the total Gasused of Precommit Sector decreased by about 60%.
2. Aggregated data is not necessarily related to Gasused. At the same time, due to the limited data samples, it is not completely possible to determine whether the more aggregated sectors, the lower the average Gasused will be. And on Github, the contribution of Precommit Sector to share sector costs through aggregation is not mentioned too much.

The assumption may not be obviously correct that Gasused is affected by the number of aggregated sectors.

Gas Fee in ProvecommitSector is reduced about 90%

Total and Average GasUsed in ProveCommitSector, from IPFSForce, 20210623

According to Provecommit Sector data provided by IPFSForce, there are two main changes:

1. The reduce of GasUsed in ProveCommitSector is more obvious, by about 90%. According to previous data, the average Gasused of the original Provecommit Sector is hundreds of millions, which was about 30 million in the test phase,by a decrease of about 90%.
2. The higher the number of aggregated sectors is, the lower the average GASUSED will be. As can be seen from the figure above, the average Gasused will be lower as the number of aggregated sectors increases.

In general, Hyperdrive has a significant contribution on fee reduction on the calibration network, especially for PoRep.

When the network is upgraded on July 1, it will surely bring a thriving state of Filecoin network with great improvement on storage efficiency.

At the same time, Hyperdrive will inevitably bring some new calculation methods of Gas fee, but the adjustment is not out of expectation. After HyperDrive, the author will share nore details on Gas fee calculation for investors.

PS: This article is only for your research and analysis. Please do not consider it as an investment recommendation.