PopDEX基于Morph Tachyon构建高性能链上交易基础设施，推动链上永续交易性能升级
Related
Analysis: Coinbase is deeply integrated with the DeFi lending protocol Morpho, with deposits accounting for over half and lending exceeding $1.2 billion.
According to Mars Finance, on June 21st, on-chain data showed that Coinbase assets have become the most important source of traffic for the decentralized lending protocol Morpho. USDC and cbBTC together account for 53% of Morpho's total platform deposits, with Coinbase users contributing 83% of cbBTC deposits and 59% of USDC deposits. The lending side is also highly dependent on the Coinbase ecosystem. Of Morpho's $2.3 billion in USDC loans, approximately $1.2 billion came from Coinbase users using cbBTC as collateral, accounting for more than half. Analysts point out that the two have formed a deep symbiotic relationship: Coinbase undertakes the functions of user acquisition and asset referral, while Morpho, as the revenue layer, brings these assets into on-chain operation. This case further confirms that in the crypto industry, the moat of distribution channels is often stronger than the technology itself.
Morph releases Agentic Economy report: Commercial GMV impacted by AI agents is projected to exceed $500 billion by 2028.
According to Foresight News , Morph released a report titled "The Agentic Economy," which predicts that by 2028, the global merchandise volume (GMV) influenced by AI agents will exceed $500 billion. The report argues that AI is evolving from a content generation and suggestion tool into an economic participant capable of discovering goods, initiating payments, and completing transactions. The report points out that the Agentic Commerce infrastructure is forming around four layers: identity, authorization, checkout, and settlement. Standards such as MCP, AP2, ACP, UCP, Visa Trusted Agent Protocol, ERC-8004, and x402 are collectively constituting the technology stack for autonomous business interactions. Morph predicts that by 2028, AI-initiated stablecoin payments will surpass human transactions in terms of the number of transactions, and one in ten American households will regularly authorize AI agents to make purchases on their behalf.
BSC releases report on quantum-resistant cryptography migration: Transaction signatures have been switched to ML-DSA-44, resulting in a 40%-50% decrease in TPS tests.
According to Mars Finance, on May 19th, BNB Chain released the "BSC Post-Quantum Cryptography Migration Report" on May 14th, stating that it has completed quantum-resistant cryptographic migration tests for transaction signatures and consensus layers, adopting the NIST-standardized post-quantum signature algorithm ML-DSA-44 (Dilithium) and the pqSTARK aggregation scheme. The report shows that BSC has replaced transaction signatures from ECDSA to ML-DSA-44 and switched consensus voting aggregation from BLS12-381 to pqSTARK to address the potential threat of future quantum computing to existing elliptic curve cryptography systems. However, post-quantum signatures also significantly increase on-chain data volume: the size of a single transaction increases from approximately 110 bytes to approximately 2.5KB; the block size increases from approximately 130KB to approximately 2MB in a 2000 TPS scenario; and TPS decreases by approximately 40%-50% in the test environment. BSC states that the current network bottleneck primarily stems from the propagation of larger transaction data, rather than the consensus protocol itself. Meanwhile, consensus layer aggregation remains highly efficient, with pqSTARK achieving a signature compression ratio of approximately 43:1, and the additional burden on validators remains manageable. The report concludes that current technology can achieve "quantum-resistant" blockchain deployments, but future challenges still include network bandwidth and data scalability.
BNB Chain plans to launch a new Layer 1 blockchain, with the mainnet expected to go live in 2027.
According to Mars Finance, BNB Chain is developing a new Layer 1 blockchain for Agentic Trading and has released its first detailed architecture information after months of development. According to BNB Chain's technology roadmap for the second half of 2026, the new chain will run in parallel with the existing BNB Chain ecosystem, aiming to achieve transaction preconfirmation of less than 50 milliseconds. This aims to provide an execution experience close to that of a centralized exchange (CEX) while retaining the advantages of on-chain self-custody and transparency. In terms of technical architecture, the new chain will remove the traditional public mempool and introduce a transaction transmission mechanism called "TxStream," which directly sends transactions to block producers to reduce latency and minimize MEV extraction behaviors such as sandwich attacks.
TermiX, the agent-based commercial clearing layer, has officially launched mainnet and released the BNB Chain ecosystem Agents Marketplace: Agent.family
According to Mars Finance, TermiX, an AI agent commercial clearing layer, has officially launched mainnet and introduced the BNB Chain ecosystem Agents Marketplace. Agents can register ERC8004-based Agent IDs on the marketplace and post, accept, and verify over 120 different types of Agent Jobs. Major agents, including OpenClaw and Hermes, can integrate TermiX Skills with a single click, gaining the ability to conduct agent commercial activities globally. TermiX is reportedly an AI agent commercial clearing layer on the BNB Chain, and in May of this year, it announced a $1 million investment from YZi Labs.
MIRA, an open-source 5B multiplayer world model, uses DINOv3 representation to mitigate long-term drift and can simulate 2v2 battles in Rocket League in real time.
According to Beating, AI research firm General Intuition, in collaboration with French AI lab Kyutai and Epic Games, has launched MIRA, a multiplayer interactive world model. As a generative game simulator supporting real-time multiplayer interaction, MIRA can simulate 2v2 battles in Rocket League in real time, based solely on historical footage and player button presses, without requiring a physics engine, rendering engine, or explicit 3D representation for inference. Unlike the "decoupling of logic computation and image rendering" approach adopted by companies like Odyssey, MIRA takes a generative simulation approach based on video latent space. MIRA boasts 5 billion parameters, and its core design builds the latent prediction space on a frozen general-purpose visual encoder, DINOv3-L. Leveraging pre-trained visual features, the generated latent states can more stably fall within the effective representation space, significantly mitigating image drift and divergence during long-term prediction. For multi-screen alignment, MIRA stitches the latent images from four player perspectives into a unified grid, enabling spatial attention mechanisms to operate naturally across viewpoints and improving the spatial consistency of vehicles, the soccer ball, and key events across multiple perspectives. The Action Dropout introduced during training also helps the system complete the game behavior of vehicles not controlled by commands when parts of the motion flow are missing. Currently, MIRA can run in real time at 20 frames per second on a single NVIDIA B200 graphics card. The team has open-sourced the training and inference code and released the Rocket Science dataset, which contains 1,000 hours of matches, approximately 4,000 hours of video, motion flow, and physics data from four perspectives; the complete training of the model used approximately 10,000 hours of clean match data.