Since the beginning of this year, Filecoin has occupied the market like a storm. Before the test is around the corner, the official is releasing news more frequently. The third miner conference call was held on 22nd July with following highlights announced:
1、The testnet will starts on August 3rd；
2、Calibration window period: July 20th-August 2nd；
3、The preliminary parameters of the economic model are being coded, but there is still a long way to go before the final version；
4、The reward is determined by storage capacity and transaction success rate；
5、The robot is set up in Europe, so Chinese miners should be well prepared for testing the network environment in advance；
6、In order to ensure the fairness of the game, the official will verify the miner address and anyone who falsifies the location will be banned and the cheater’s rewards will be cancelled。
The official revealed the criteria for testnet awarding
Although the economic model has not been confirmed, Filecoin's development direction has actually already appeared after many online Q&A. For FIL, the economic model is not only data planning, but more importantly, it determines various rules of Filecoin. I believe that we can have an accurate answer before the mainnet goes live. At that time, RRMine will conduct a deeper interpretation of the economic model
In addition, this the miners meeting also revealed another important thing: what is the decisive factor for the testnet contest? According to the official statement, the reward criteria are original storage capacity, storage success rate and sector update cycle. On the basis of the original 4 million Filecoin incentives, an additional 100,000 Filecoin will be allocated to the top 20 block reward winners. In addition to increasing storage capacity, miners must achieve a certain success rate (pass/fail rate) of real storage transactions that is higher than a standard if miners want more incentives (details of the standard to be determined).
The core concepts of Filecoin consensus and storage algorithms
From the perspective of the ranking, the original storage capacity, storage success rate and sector update cycle are the decisive factors. Therefore, let’s consider how can we get more orders and greater opportunities for block generation from the perspective of the core concepts of Filecoin consensus and storage algorithms.
Unlike Bitcoin whose block generation is decided by hash power, the core of Filecoin generation is storage capacity, whose core concept is "Expected consensus + Proof-of-Replication + Proof-of-Spacetime". The expected consensus determines what method and who are selected to win the block award; the more data that Filecoin miners effectively store, the greater the possibility of block generation.
In addition, Filecoin has two storage algorithms to verify storage capabilities: Proof-of-Replication (PoRep) is used to confirm that the miner has successfully replicated and encapsulated the data in its own device and Proof-of-Spacetime (PoSt) is used to verify that the miners store data constantly and safely as promised.
How to accumulate more "storage capacity" to obtain greater opportunities for block generation? The first is early participation. Filecoin and bitcoin mining are like a marathon and a 100-meter race. FIL mining needs "accumulation", which means early participants can not only get the "early mining bonus", but more importantly, they can accumulate effective storage at the very beginning. The first-mover advantage is to accumulate storage capacity, while latecomers need to pay much more money to make up for the disadvantages comparing with the first movers.
In addition to early accumulation, there is another way, that is, to increase the growth rate of storage capacity, which needs faster sealing speed and stable hardware to store valid data that has been verified, so as to achieve the rapid growth and the later comers surpass the formers.
Worker + Miner solution, multi-threaded concurrent processing
The IPFS storage of RRMine is connected to Athena mining pool by software and hardware cluster mining solution and supercomputing cloud service + mining rigs. Compared with all-in-one products which means the entire mining process is completed by one machine, the cluster construction makes full use of the advantages of CUP, GPU, hardware I/O, bandwidth, etc., maximizes hardware utilization and ensures higher outputs per unit time and cost.
The technical requirement of cluster construction is high, which requires a deep understanding of Filecoin mining before the optimal solution is applied. RRMine adopts the technical solution of Worker+Miner. The Worker is responsible for processing various data from mining, including all data of zero-knowledge proof (Proof-of-Replication and Proof-of-Spacetime). The Miner is responsible for receiving orders, synchronizing messages and coordinating tasks. The coding tasks of multiple workers can be processed efficiently at the same time with such multi-threaded concurrent processing. Compared with the single-threaded tasks of other products, the unit cost and efficiency is twice.
Follow the general direction and seize the growth chance of Filecoin
The advent of digital economy age has given IPFS/Filecoin an opportunity to climb. IPFS/Filecoin, together with cloud storage, cloud computing, Internet, big data, etc. will form the "new infrastructure" of WEB3.0 and prompt a new blue ocean market. Ethereum moved its official website to IPFS; Huawei built an IPFS storage centre in Cangzhou; JD Cloud is directly linked with IPFS! The trend of these giants is the general direction of the future.
Computing power itself is a high-quality asset and also the most innovative core productivity as the "new energy" in the world of digital economy. RRMine will firmly take advantage of the general trend of computing power platformization after three years of experience accumulation. Driven by technological innovation and empowered by ecological operation and maintenance, we will face the challenges and opportunities brought by the rapid growth of information technology of new infrastructure.
