TBW2025》Google and AMD emphasize that "AI + Web3" is crucial: chip computing needs Blockchain to upgrade trust.

At Taipei Blockchain Week (TBW), executives from Google Cloud and AMD revealed the blueprint for the convergence of AI and Web3, and emphasized that "confidential computing" and blockchain trust are key technologies to build the next generation of trusted digital ecosystem. (Synopsis: AI eats too much electricity!) U.S. restarts nuclear power plant Duane Arnold: Response to data center in Iowa) (Background added: Jensen Huang shouts "no cryptocurrency"!) Nvidia suddenly shouted to stop Arbitrum cooperation, why did her attitude take a big turn? Artificial intelligence (AI) and Web3 are undoubtedly the two most turbulent waves, but when people mention the combination of the two, many people question whether this is just a mixed hype of the noun, but on the stage of the 9/4 Taipei Blockchain Week (TBW), Jörg Roskowetz, head of AI and Web3 strategy of chip manufacturer AMD, and Eefy Lin, a leader of Google Cloud Web3 architecture engineer, told the audience that AI and Web3 Integration will be the core key of the new generation, drawing a clear blueprint for integration for the industry, and the core message of the dialogue is straightforward and powerful: "Confidential Computing", which has been built from the chip level, will no longer be an optional option, but an inevitability, and blockchain is an important cornerstone for building a future trust ecology. AI meets blockchain: "trust" as a technology giant hinders Jörg, an "OG" who has been in the Web3 field for eight years, even at four o'clock in the morning in Germany, enthusiastically explained how AMD starts from the lowest hardware and lays a security foundation for this new era through remote enthusiasm. Subsequently, Eefy Lin revealed from the perspective of platform and ecology, how Google Cloud has not only become the preferred platform for Web3 developers through a comprehensive strategic layout, but even began to build its own "common ledger", pointing to the core of the financial industry. Blockchain is the most transformational technology of the last 50 years, and now we have added a new component, artificial intelligence, which is no longer a stand-alone technology, but works hand in hand with blockchain as a building block. This view shatters many people's conventional wisdom about both. If blockchain is like an immutable "Truth Machine," then AI is the "Decision Engine" on which it is built. However, a key question arises: when we feed massive amounts of data to AI models for training and inference, how do we ensure the privacy and security of this data during processing? How can you trust that AI's decisions are not being manipulated maliciously? Jörg cites data that the amount of data stored in the cloud worldwide is growing at a staggering rate of 25% per year and is expected to reach 527 zettabytes by 2029 (1 zettabyte equals 1 billion terabytes). The explosive growth of data means an infinitely expanding attack surface. Especially in highly sensitive fields such as finance and healthcare, data security is a top priority. We expect 575.1 billion digital payment transactions in 2028. We need to ensure that all these different payment methods are protected in a confidential manner. Similarly, when it comes to life-saving health data, I believe none of us want our own health data and DNA data leaked to third parties. This brings us to AMD's core solution: confidential computing. When AI models and personal privacy data circulate in the cloud server, how to prevent cloud service providers, malicious insiders and even hackers from prying eyes? Traditional at-rest encryption (at-rest) and in-transit encryption (in-transit) are no longer enough, and the risk of data leakage in use (in-use) has become the biggest security gap. AMD's Hardcore Solution: "Confidential Computing" Creates Security Islands from the Chip Layer Faced with security challenges in data usage, Jörg proposed AMD's "hardcore" solution. Confidential computing, he explains, is like building a separate and encrypted "secure enclave" inside the CPU. Any data and application code that enters this silo is encrypted at the hardware level, so that even the most privileged system administrators or the cloud platform itself cannot peek into its inner workings. "We achieved this with AMD Infinity Guard technology," says Jörg. This technology includes key features such as Secure Cryptographic Virtualization (SEV) and Secure Nested Paging (SNP) to create an isolated execution environment at the virtual machine level. The latest advance is Trusted Domain Isolation and Security Configuration (TTISP), a technology that extends the trust boundary from CPUs to entire computing platforms such as storage, smart NICs, and AI accelerators, ensuring that data is encrypted throughout the end-to-end process. It's worth noting that AMD isn't building cars behind closed doors. Jörg repeatedly emphasized the importance of community and open standards, which is why AMD co-founded the Confidential Compute Consortium. This alliance brings together Google, Microsoft, Intel and many other industry giants to jointly promote the standardization of related technologies. "The community is the strongest ally of all technologies," says Jörg, "and our philosophy is open standards, because that's the strongest backing we have." We work closely with trusted partners like Google Cloud to integrate these standards into the final solution." More than just selling hash power! Google Cloud's Web3 full strategy After the wonderful computing power island speech, Eefy Lin revealed how Google Cloud has not only become the preferred platform for Web3 developers through a comprehensive strategic layout from the perspective of platform and ecology, but even began to build its own "universal ledger", pointing to the core of the financial industry. If AMD has built a "foundation" of trust from the hardware level, then Google Cloud has built a "highway" to the mass adoption of Web3 on top of this. In his presentation, Google Cloud's Eefy Lin elaborated on Google's panoramic Web3 strategy, which is ambitious to go far beyond providing infrastructure. We want Web3 to shine, we want to be the platform of choice for all Web3 participants, make blockchain data easy to analyze, and make the infrastructure itself a trusted layer. To achieve this vision, Google has adopted a multi-pronged strategy. The first is to inject more than $1.5 billion into the Web3 field through Capital G and GV, the investment institutions of parent company Alphabet, directly supporting the innovation force in the ecosystem. Second, Google is integrating Web3 features into its vast product matrix: the Google Play Store provides a channel for the distribution of Web3 apps; Google Wallet and Google Pay are also working on integrating cutting-edge technologies such as zero-knowledge proofs (ZK) in an attempt to seamlessly bring billions of users into the Web3 world. For developers, Google Cloud offers a powerful arsenal. Its BigQuery service transforms complex on-chain data into easy-to-analyze insights, freeing developers from having to parse the underlying data. In addition, the node serves...