module

Developers discovered that after deconstructing the official installation package of Xiaomi MiMo Code, it contains two modules, trajectory-bundle and codebase-bundle, which are not available in the public repository. The former can package system prompts, dialogues, model responses, and code diffs; the latter includes collectCodebase(), which can traverse Git repositories, read source code, and compress packages.Currently, there is no evidence that collectCodebase() is called by the built-in code of MiMo Code, nor is there any evidence showing that the complete codebase has been uploaded. The external content confirmed by developers mainly includes information such as Git repository addresses, commit hashes, and branches. Xiaomi MiMo Code was open-sourced under the MIT license in June.Two days ago, ZhiPu ZCode faced a 12-page legal letter from a company demanding the deletion of data due to automatic packaging of the complete workspace.

According to Bitcoin News, Flipper's upcoming LoRa module can send messages through a radio relay network without relying on the internet. In tests conducted in London, messages were transmitted over 150 kilometers via the MeshCore network.This technology can be used to send Bitcoin transactions: signed Bitcoin transaction data can be transmitted via LoRa radio and relayed to a gateway connected to Bitcoin nodes, which then broadcasts it to the network. When local internet connections are interrupted, users can send transactions in this way.

Micron showcases the world's first 512GB DDR5 RDIMM memory module designed for servers, with a maximum speed of 9200 MT/s. AMD and Intel are validating it for future platforms. This module achieves capacity through silicon via vertical stacking of DRAM chips, and a server memory configuration with 24 modules can reach up to 12 TB.The single 512GB module has a rated power consumption of 16W, while four 128GB modules have a total power consumption of 44.2W, a reduction of over 60%. Micron claims that in memory-intensive analysis scenarios using Spark SVM, performance is improved by up to 1.4 times compared to 256GB DDR5. This product continues the 256GB DDR5 RDIMM that Micron sampled earlier this year (also with a maximum of 9200 MT/s, using 1-gamma DRAM and 3D stacking), with mass production planned for 2027, and pricing has not yet been announced.

Micron Technology (MU.O) stated that the company has achieved a milestone with the 512GB DDR5 memory module and expects that the 512GB DDR5 RDIMM will enter mass production in the second half of 2027.

Goldman Sachs significantly raised its industry forecast in the latest global optical module report, expecting the global optical module market size to reach approximately $67.7 billion, $131.4 billion, and $148.5 billion in 2026, 2027, and 2028 respectively, an increase of 33%, 81%, and 115% compared to previous forecasts.Goldman Sachs believes that this round of upward revision is not solely due to the increase in the number of AI servers; more importantly, the architecture of AI servers is continuously migrating towards high-speed networks, coupled with the increase in the number of optical modules required for a single GPU or ASIC chip, driving rapid growth in demand for 800G, 1.6T, and even 3.2T.

NVIDIA (NVDA.O) announced that the Jetson Orin Nano 2 module and kit are expected to be launched in the first half of 2027.

Cosmos Labs announced that an ongoing security incident has affected users of the Cosmos EVM module. The security and engineering team at Cosmos Labs is actively responding to the incident and has advised validators on the Cosmos EVM chain in contact with them to suspend the operation of their chain.Cosmos Labs appreciates the swift response from all teams using Cosmos EVM and stated that an incident report will be released after the situation is resolved. If relevant teams are using Cosmos EVM and have questions, they can contact security@cosmoslabs.io.

According to the Industrial and Commercial Times, driven by the long-term growth in demand for memory and advanced packaging due to AI, Winbond Electronics has proactively initiated expansion plans for its Kaohsiung Luzhu plant, merging the originally planned phases two and three into Module B for simultaneous development. Construction of the cleanroom is expected to start in 2027, with equipment installation beginning as early as the beginning of 2029, and equipment will be phased in according to customer demand forecasts and LTA. Currently, there are already customers negotiating for capacity in 2029 and 2030.Industry sources indicate that Winbond's niche DRAM and SLC NAND are in short supply in the third quarter, with a quarterly price increase estimated at about 50%; NOR Flash benefits from large cloud service provider customers stocking up, with a quarterly price increase estimated at about 30%. The price increase for memory in the fourth quarter is expected to converge to 2% to 5%, but thanks to increased DRAM capacity and growing shipment volumes, revenue and profits can still maintain a quarterly increase. Winbond's consolidated revenue in the second quarter was NT$59.843 billion, a quarterly increase of 56.4% and a year-on-year increase of 184.7%. DRAM prices increased by about 100% quarter-on-quarter, while Flash rose by about 43%, with consolidated gross margin rising to 66.2% and after-tax earnings per share of NT$5.40.The Module B product plan includes Standard DRAM, CUBE DRAM, Wafer-on-Wafer (WoW), and silicon capacitors (Si-Cap), which will support 14-nanometer and future 12-nanometer DRAM processes, with plans to introduce EUV equipment in the future.

According to the "Science and Technology Innovation Board Daily," as the original factory prices further rise in the third quarter, the price of server DDR5 has increased by 15% to 23% this month, with 32GB / 64GB / 96GB DDR5 RDIMM rising to 900/1590/2650 USD respectively. In the spot market for server RDIMM, the trade price for 32GB D5 is between 1800~2100 USD, 64GB D5 is between 3000~3200 USD, and 96GB D5 is between 3800~4100 USD. The price difference between contract and spot DDR5, especially for 96GB, shows a trend of narrowing. The supply of original factory DDR5 remains insufficient to meet demand; however, as contract prices rise seasonally, large server clients are becoming more sensitive to the increase in memory costs. Additionally, with long-term agreements being established, the supply of D5 for large clients is somewhat guaranteed, and the increase in DDR5 prices is expected to be more moderate.

According to official news from the decentralized stablecoin USDD, the DeFi protocol Pendle has integrated PT-s USDD into its PT Looping module. Users can complete looping loan operations with one click through the Pendle interface, without the need for cross-platform operations. Currently, the looping loan APY for PT-s USDD in the Morpho/USDT market can reach up to 20.87%.It is introduced that the PT Looping module is based on the PT-s USDD/USDT and PT-s USDD/USDC markets on Morpho. This integration aims to significantly enhance the capital efficiency for USDD holders and, combined with the fixed income of PT-s USDD and borrowing space, forms a robust yield portfolio. The official reminds users that they can participate according to their own risk preferences.

According to Counterpoint Research analysis, if the proposed import ban on Chinese optical modules by the Federal Communications Commission (FCC) in the United States is implemented, it will primarily impact American cloud providers rather than Chinese suppliers. InnoLight leads the global data center optical module revenue with approximately 27% market share, while Coherent ranks second with about 17%. Chinese manufacturers collectively account for about two-thirds of the global unit supply and approximately 60% of the optical communication module revenue.The analysis points out that Western suppliers like Coherent and Lumentum lack sufficient cleanroom capacity, automated packaging infrastructure, and yield scale in the short term, making it impossible to fill the capacity gap left by Chinese manufacturers within 12-24 months. This could lead to delays in AI cluster deployments by several quarters and increase material costs for cloud providers. At the same time, Chinese module manufacturers rely on high-end 800G/1.6T module revenue from North American hyperscale customers for over 90%, but have achieved partial capacity relocation by establishing factories in Thailand.Counterpoint emphasizes that the optical module supply chain is a highly interdependent system. Chinese manufacturers heavily procure DSP chips from Broadcom and Marvell, as well as lasers and optical chips from Lumentum, Coherent, and Mitsubishi Electric. A forced separation would disrupt the entire ecosystem.

According to Reuters, citing four informed sources, the Trump administration is drafting a ban on the import of new models of optical modules for data centers from China. The Federal Communications Commission (FCC) is formulating this measure to prohibit the import of new models of optical modules that can transmit data through fiber optic cables within data centers, with officials hoping to release and implement it within this year. The ban aims to prevent Chinese companies from stealing data, implanting malware, or disrupting services at U.S. data centers.The ban may primarily affect Inspur (leading with a 27% share of the global data center optical module market), which was listed in June of this year on the Pentagon's list of companies involved in military activities. U.S. cloud providers (such as AWS) may be forced to turn to alternative suppliers like Coherent and Lumentum, raising costs. The FCC has previously imposed similar restrictions on Chinese drones, routers, robots, and inverters. The Chinese embassy in the U.S. stated that "it will take all necessary measures to respond to any actions that harm China's interests." The draft may still be modified or shelved.

According to Blockaid monitoring, the trading price of MetronomeDAO's Synth USD (msUSD) on Ethereum, Base, and Optimism once dropped about 11% below the pegged value. Subsequently, MetronomeDAO released a post-mortem report stating that its synthetic asset swap module had insufficient collateral issues, resulting in a lack of adequate asset backing for approximately 6,367 msETH and 4.57 million msUSD, with the impact primarily concentrated on that module.The report pointed out that the problem stemmed from delays in the Chainlink price oracle during swap execution, and Metronome's fee design failed to adequately account for this variable, particularly on the Base chain where the situation was more severe. The team has deployed over $34 million in defensive positions and about $6.5 million in "last exit" liquidity, which could be used to repurchase and destroy sufficient assets to eliminate the gap if the peg deviates by about 30%. The team has increased the transaction fees for all synthetic trading pairs and upgraded the protocol to support a directionally isolated fee mechanism. Priority will be given to ensuring the treasury repurchases and destroys synthetic assets to restore 100% adequate collateral, while not affecting the rights of MET token holders, and the MET repurchase and distribution plan will proceed as usual.

The Web3 asset platform RootData has now launched the RD AI Popularity Interpretation module. This feature uses a specific AI model to deeply analyze the reasons behind project popularity, helping users quickly grasp the latest dynamics and core driving factors of projects when viewing the homepage's popular project rankings, thus enabling more efficient identification of project value and improving investment decision-making efficiency.RootData has previously launched core indicators such as the RD Popularity Index and RD Growth Index, which present real-time rankings of the top 100 projects by popularity index. The newly added AI interpretation feature will further enhance the platform's data insight capabilities, providing users with more intelligent tools to discover early alpha projects and investment opportunities.

According to the official announcement, Bitget has made multiple optimizations to its CFD copy trading feature. This upgrade focuses on two main scenarios: "Expert Discovery" and "Position Management," aiming to help users efficiently filter high-quality trading experts and manage copy trading positions more clearly. Users can upgrade the Bitget App to version 2.87.0 to experience the related features.In terms of Expert Discovery, the copy trading tab on the market K-line page has added a "Popular Trading Experts" module. Users can view the popular trading experts with the best performance over the past 30 days for the current asset. The expert card will display information such as nickname, 30-day return rate, and the number of copy traders. Users can directly enter the expert's homepage to view strategy details or initiate copy trading with one click, reducing the cost of switching between the market page and the copy trading center.Regarding Position Management, the CFD trading area has added a "Copy Trading Position" module, allowing users to directly view the copy trading positions under the current asset in the position column. If following multiple trading experts simultaneously, the positions of each expert will be displayed independently in separate lines and can be sorted by position value, unrealized profit and loss, and other dimensions, helping users manage their copy trading accounts and risk exposure more precisely.

The Zodiac team released a security incident analysis report regarding the impact on the Zodiac Roles Modifier, disclosing that the root cause of the vulnerability lies in a flaw in the ERC-1271 transaction signature verification logic: the system only determines the validity of the signature based on the returned "magic value" without verifying whether the call itself was successful, which may disguise a failed verification as a valid signature, bypassing the module authentication mechanism.

According to The Block, Gnosis co-founder and CEO Martin Koppelmann confirmed that the Zodiac delay module related to Gnosis Pay is currently under exploitation. Attackers can initiate transactions from Safe wallets that integrate this module, and Gnosis has requested cross-chain bridge validators to pause to control the risk.Koppelmann stated that Gnosis will cover all user losses and will remove the previous announcement urging users to urgently withdraw EURe and GNO, noting that most users are unable to withdraw their tokens independently. The team is working to control most of the losses and ensure that all users receive full compensation. Gnosis emphasized that the vulnerability is within the Gnosis Pay system, and the core contracts of Safe are unaffected.

The founder of Slow Fog, Yu Xian, posted on the X platform to interpret the Squid security incident. He stated that sampling found that the related Safe wallets were all single-signature, and the owners were all different, but the issue was not with the private keys; the problem lay in the modules (SquidRouterModule) used by these Safe addresses, which had vulnerabilities. Attackers could forge messages, easily bypass relevant validations, and initiate subsequent exchange operations to transfer funds from the target Safe wallet. In addition, Yu Xian also disclosed information about the address where the attackers' profits were deposited.Previously, it was reported that a third-party Gnosis Safe module was exploited on Base and Ethereum, resulting in a loss of approximately $3.2 million, with the victims being 86 Gnosis Safes that had added this contract as a trusted Safe Module. The contract was named "SquidRouterModule" on Basescan, and subsequently, Squid clarified that it was not affected by the Gnosis Safe-related vulnerability incident.

The cross-chain routing protocol Squid stated on platform X that the SquidRouterModule involved in the attack disclosed by Blockaid was neither developed, deployed, nor operated by Squid, but rather a third-party Gnosis Safe module that chose to integrate with protocols like Squid, with no prior connection between the two parties.Squid indicated that this third-party module has a vulnerability that accepts a fixed string provided by the caller as message security verification. The attacker exploited this string, which is publicly visible in the verified code of the contract, to execute arbitrary call data and steal funds. Squid's own routing contract (0xce16F) is architecturally completely different from this module and is unaffected; user funds, authorizations, and integrations are all secure.

Aave announced that its bug bounty program has been updated to better align rewards with the risk status of each part of the ecosystem and to simplify the review process. The maximum reward for critical vulnerability fixes for Aave V4 and Core Aave V3 has now been increased fivefold.