Mining machines turned into scrap metal, a history of millions of blood and tears Do you still remember FIL's legendary 'King of Storage'? I bought a mining machine for 70000 to 80000 yuan, but the CX boss tricked me into saying 'three generations of descendants will have enough to eat'. What was the result? The pile now is full of scrap metal with QR codes, the electricity bill has not been recovered, and all the capital has been fed to the dog! In 2022, I had dinner with several big brothers who make Filcoins in Shenzhen. They really put all their wealth into it. At first, the cost was nearly 20 yuan, but gradually decreased to 10 yuan, and the result was still a loss with no underwear left. An old man told me that Filecoin has attracted a large number of people in Shenzhen, some of whom even mortgaged their houses and ended up losing everything 🧵 On the contrary, a friend recently showed me a hot storage solution developed by Shelby @ shelbyservants Jump and Aptos in collaboration. Yes, it's another storage solution Firstly, technically speaking, Shelby focuses on real-time data interaction capabilities, with read and write latency in the sub second range. Unlike Filecoin's cold storage, the read speed is extremely slow. And its economic model is also quite interesting, as data can generate revenue every time it is read, and creators can directly set charging rules. If you haven't tried using Fil and AR storage, you may not know that manual retrieval is very slow. If you really want to use storage as a server, it must support millisecond level response time. Although Filecoin and hot storage are positioned differently, they can form complementary relationships Long term archiving of files (such as historical data, NFT metadata) can be stored on Filecoin. Real time interactive data (such as dynamic updates, AI model real-time feeding) can be stored in thermal storage systems (such as Shelby) Filecoin focuses on low-cost decentralized storage, suitable for static data, while hot storage focuses on high-frequency interaction, suitable for dynamic data. It is not said which one will be better, but for different usage scenarios. Currently, there are more dynamic scenarios that need to be supported, and Shelby has already partnered with big brands such as Story protocol and NBC universal Behind Shelby are Jump and Aptos scenes. The transaction confirmation speed of Aptos chain is unbelievably fast, it can be completed in 600ms, and the cost of a single transaction is only $0.00055. Jump has brought an institutional level throughput architecture, which ensures Shelby's performance To sum up, what about AR for permanent storage, STORJ for enterprise services, and FIL? I am still teaching people how to start a room. I can only say that this industry is indeed evolving, with new stars emerging in the storage industry. There are always better solutions and technological solutions available