Why do I feel that FHE hasn't reached its explosive moment yet? Recently, I've read a lot of articles about FHE (Fully Homomorphic Encryption). When Multicoin led a $73 million investment in Zama, they boldly claimed: "ZKP is in the past, the future belongs to FHE." To be honest, I was shocked by this statement. But after the shock, I calmed down and thought: Is FHE really ready? My answer is: not yet. Why did ZK explode? Because there is an economic demand. Let me first explain why ZK became popular. Many people think ZK became popular because of "privacy protection," but that's not the fundamental reason. The fundamental reason ZK became popular is: it can help Ethereum save money. Ethereum needs to scale and has established a Layer 2 route, while ZK Rollup can compress a large amount of transaction data into a small proof and send it back to L1. This has enormous economic value. Thus, the four giants ARB, OP, zkSync, and StarkNet have risen, with ZK hardware acceleration, ZK programming languages, ZK modularization... the entire ecosystem is crazily competing. This is not driven by technology; it is driven by economic demand. Only after the demand for "L2 needs to efficiently transmit data to L1" emerged did ZK's implementation get triggered. It is not a case of "we have ZK technology, let's find a scenario to use it." This is like looking for nails with a hammer, forcing a match, which will only increase the implementation cost. The problem with FHE: it hasn't found a must-have scenario. Now back to FHE. The technology of FHE is impressive; I do not deny that. It allows encrypted data to participate in computation directly without needing to decrypt. This sounds cool, but the problem is: who needs this? I've looked at many FHE project whitepapers, and the application scenarios they list are nothing more than: Privacy transactions Resistance to MEV Safer networks Preventing third-party snooping But none of these are first-order needs. Why? Because Web3 users actually don't care that much about privacy. Only when privacy can provide economic value will users use related tools. For example, hackers use Tornado Cash to hide stolen funds. But ordinary users will use Uniswap because Tornado Cash incurs additional time and economic costs. The encryption cost of FHE itself further tortures the weak operational efficiency on-chain. Only when this increased cost brings more significant benefits will privacy protection have the potential for large-scale promotion. Computational efficiency: the fatal flaw of FHE. Another more practical issue: FHE is too slow. Ciphertext computation may be more than 10,000 to 1,000,000 times slower than plaintext computation. Even with the most advanced TFHE solution, Zama's current TPS is only around 5, while Inco can achieve 10 TPS. What does 10 TPS mean? The TPS of Ethereum's mainnet is around 15, while Solana's TPS is 65,000. The speed of FHE chains is even slower than Ethereum's mainnet. Of course, Inco claims that they can use FPGA hardware acceleration to boost TPS to 100-1,000. But that's still too slow. Moreover, hardware acceleration requires money, time, and ecosystem support. ZK hardware acceleration was nurtured by a booming market demand. But what about FHE? Where is the market demand? The lesson from Elusiv: being early is not as good as being timely. I saw a case that made me more convinced that FHE isn't ready yet. Elusiv was a "dark pool" protocol on Solana that started in 2022, using FHE technology. But in 2024, Elusiv ceased operations, and its codebase and documentation were deleted. Later, they rebranded as Arcium, secured new funding, and transformed into a "parallel FHE" solution. What does this indicate? It indicates that a pure FHE privacy solution is not appealing to the market. Users are unwilling to pay extra costs for privacy. Only when FHE can reduce costs or improve efficiency will the market spontaneously promote it. Who needs FHE? Institutions, not retail investors. So, does FHE have no value? No. The value of FHE lies in institutional-level applications, not retail investors. For example, in the direction of RWA, bond issuance and trading. In June 2023, China International Bank, through UBS, issued "blockchain digital structured notes" to clients in the Asia-Pacific region. But strangely, the contract address and distribution address for that transaction cannot be found. Why? Because institutional clients have a need to use blockchain but do not want to disclose all information. At this time, the ciphertext display and direct transaction features of FHE are more suitable than ZK. But the question is: how big is the institutional market? The institutional market is certainly large, but it is not the retail market. The retail market is driven by hype, FOMO, and the myth of getting rich. The institutional market is driven by compliance, stability, and long-term value. FHE may find its place in the institutional market, but it will not create a frenzy in the retail market like ZK. My judgment: FHE is still looking for its moment. I've been thinking lately: when will FHE explode? My answer is: when it finds a scenario where "FHE must be used, and no other solution can be used." ZK found this scenario: L2 data compression. MPC found this scenario: multi-signature solutions for AA wallets. But FHE hasn't found it yet. Current FHE projects are all doing the same thing: building infrastructure and waiting for scenarios to emerge. Zama is working on fhEVM, allowing developers to deploy FHE contracts on EVM-compatible chains. Inco is creating a "general privacy computing layer" to provide FHE computing outsourcing services. Fhenix is developing FHE L2, hoping to become part of Ethereum. These are all good explorations, but none have found a killer application. I'm not saying FHE has no future; I'm just saying: FHE's moment hasn't arrived yet. It needs to wait for a scenario, a "both-and" scenario: Both need to be open and transparent (the characteristics of blockchain) And need privacy protection (the characteristics of FHE) And this scenario's economic value must be large enough for users to be willing to pay the extra cost for FHE. When this scenario appears, FHE will truly explode. My action I will continue to pay attention to FHE, but I won't go all in. Because I know that the maturity of technology and the maturity of the market are two different things. The technology of FHE has matured, from Gentry's ideal lattice algorithm proposed in 2009 to the current TFHE solution, the theory is already complete. But the market is not ready yet. The market needs a story, a story that can make retail investors FOMO. And this story, FHE hasn't told yet. So, I will wait. Wait for FHE to find its story, wait for the market to be ready to accept FHE. Only then will it be FHE's true moment. @zama_fhe #ZamaCreatorProgram