As the mainnet launch date gets closer, the outside world's enthusiasm for Filecoin is also rising, but at the same time, many issues have also begun to be discussed in the 1475 community, such as:

1. The current total network computing power is 20PB. What is the effective computing power after the main network is launched?

2. The big miner test has been postponed. When will the main network be launched?

3. Market chaos occurs frequently, how can investors avoid pitfalls?

4. With the coexistence of three types of mining machines, what will happen to Filecoin mining machines in the future?

Then Andy answered all the questions at once.

On July 11, Filecoin core developer Why claimed to have received messages from many people asking for more time to prepare the "testnet incentive plan." Therefore, he initiated a discussion to hear the community's voice: How many people want an extension?

On the engineering side, the Filecoin team has been working hard to launch on July 20th, and it is possible to launch on time. But Why is also considering, perhaps delaying it by 1-2 weeks and running another devnet for another week.

What does Andy think about the news of the postponement of the big miner test?

Filecoin hopes to build a decentralized storage network through blockchain to store the most important information for human society. Filecoin itself has many breakthrough innovations, and the system development and economic mechanism design require repeated thinking and long-term polishing. And we have seen that many traditional large-scale Internet projects are often delayed. Therefore, in my opinion, the delay of Filecoin, on the contrary, shows that Protocol Labs attaches great importance to the functional realization and safe and stable operation of the project. The project party is truly working hard to move forward in the original direction instead of compromising with investors and the market.

Officials say Filecoin will be launched between the end of July and the end of August, but many people are skeptical about this. At the Filecoin Global Miners Conference, the official decided to continue using SDR as the underlying algorithm for Filecoin, and the underlying algorithm has been basically determined from a technical point of view. The network will be fine-tuned and optimized in the future, and then testing will be started and the economic model will be optimized. I think unless there are major problems again, Filecoin will not be delayed for a long time, and it is likely to be launched in August.

There are already more than a hundred service providers on the market. What kind of mining machine should we choose to avoid pitfalls and pitfalls?

1. Hard disk capacity is not equal to effective storage. Currently, most mining machines on the market are priced by hard disk storage capacity (price per terabyte). Filecoin is mined by effective storage rather than hard disk capacity, and hard disk capacity is not equal to effective storage.

2. Pay attention to the input-output ratio. Don’t be overly superstitious about the test network ranking. The ranking itself has certain meaning and value, but it cannot represent the final input-output ratio. 1475 has been paying attention to the optimization of the input-output ratio. There will be fluctuations in the ranking at different stages, and the number of machines of each company is also different.

3. Pay attention to the technical strength of mining machine manufacturers. It should be noted that the algorithms and mechanisms related to the Filecoin network may change, and mining machine manufacturers should have the ability to cope with code and mechanism changes to continuously optimize and upgrade the mining machine hardware and software to ensure the best mining performance of the mining machine.

According to the 1475 block browser, the total network computing power has exceeded 20PB, and this is only the restart stage of the second test. What will the total network computing power of Filecoin reach after the main network is launched?

Since the test network has just been reset, the storage capacity is still relatively small. As for the storage capacity after the Filecoin main network is launched, it is predicted that the main network will be relatively stable for 3-5 months after launch, and the storage capacity may reach 300P in 20 days.

Andy analyzed that the official setting of Filecoin is 100P. In fact, with the current access volume of mining machines in the entire market, 100P can be completed in 20 days. After the main network is launched, more mining machines will participate in it, so there will be a 1.2-1.5 times increase. From the supply chain perspective, there will not be a sudden large-scale access of mining machines after the main network is launched, so it will be relatively stable for 3-5 months. The storage capacity may reach 300PB in 20 days. It can be inferred that it should be around 15-20PB per day.

It can be seen that in the early stage of the mainnet launch, the growth rate of effective computing power has become a key factor. Regarding how to speed up the storage and packaging of effective data, 1475 split the underlying code from the very beginning, and made more than 60 optimization points at most. Each optimization can improve the performance by 5%-10%. During the optimization process, we also combined the optimizations to improve efficiency, including outsourcing zero-knowledge proofs, and reduced the consumption of the entire computing resources in this way.

The reason why we can achieve technological leadership is that we laid the foundation during the first test period. During the first test period, 1475's Filecoin block output ranked first, and assisted the official in testing and maintaining the Filecoin network many times, helped the official discover many important bugs, and assisted the official in upgrading the network. It is the investment in Filecoin that has brought 1475 mining machines to the leading position.

The market price of Filecoin mining machines is generally divided into three levels: high-end mining machines of 200,000-300,000 yuan, mid-end mining machines of 60,000-130,000 yuan, and low-end mining machines of less than 60,000 yuan. What will be the future prices of these three types of mining machines?

Regarding the price of Filecoin mining machines, Andy said that the price of mining machines will definitely drop, first of all due to competition, and secondly due to the rapid iteration of professional mining machines. The current mining machines are just a simple combination of hardware, but these hardware combinations do not play a big role. Various optimizations must be made in the middle, including optimization of the supply chain and optimization of the hardware structure, so that the same thing can be done at a lower cost.

In addition, it is wrong to calculate the price per T, because a mining machine can be installed with many hard disks. The more hard disks are stacked, the lower the cost will be. Now the cost of hard disks is less than 200 yuan per T. Therefore, we cannot only look at the storage hard disk capacity. We still have to look at it through the overall solution of Filecoin. Calculating how many coins are produced per T and how much it costs is itself a false proposition.