width

On September 29, Shayon Sengupta, a partner at the cryptocurrency investment firm Multicoin Capital, stated that Multicoin has invested in the decentralized bandwidth network Grass through its hedge fund and venture capital fund, referring to it as the reading layer of machine intelligence. Over the past 18 months, Grass has recruited contributors to share idle bandwidth through crypto incentives for querying and parsing the public internet, connecting data to the data pipeline of frontier labs.The article cites a report from Grass itself, stating that over 6 million contributors have received rewards denominated in USD, funded by the revenue generated directly from their contributions. Grass reported revenues of $17 million in 2025, with another $17 million recorded in the first half of 2026, with nearly all AI clients making repeat purchases. Grass claims the business is profitable and projects that revenue from training data alone will reach $75 million in 2026, with relevant figures coming from its call on July 7 to token holders and network participants.Sengupta wrote that Grass plans to launch a Contents API, allowing models to open and render public pages through residential networks, and plans to introduce a Search API along with web indexing. Grass has collected over 1 billion public internet videos, some of which are used with Multicoin portfolio company Inference.net for training video annotation models.

In episode 166 of the BidClub podcast, chip designer Bubble boi stated that high-bandwidth flash is technically attractive but lacks decisive data center use cases. The information he provided indicated that the main customers he heard about are Chinese clients who find it difficult to obtain or produce high-end HBM.Bubble boi mentioned that HBF and high-end AI flash have a latency of about 5 microseconds, lower bandwidth than HBM, a lower cost per bit, but the total cost may be comparable. He believes the clearest customers are entities unable to obtain HBM, which aligns with the Chinese demand he heard from two to three people, as they cannot adequately manufacture accelerators or are willing to accept non-leading specification products.

According to official news, Marvell Technology announced the launch of a new generation of memory solution portfolio for AI infrastructure, covering server-level AI storage, rack-level CXL memory expansion and pooling, as well as multi-rack optical interconnect shared memory, aimed at addressing the growing memory capacity and bandwidth bottlenecks in Agentic AI inference processes.Marvell stated that as AI model sizes increase, context windows extend, and KV Cache demand grows, traditional tightly coupled architectures of computing and memory are limiting AI inference efficiency. Through memory disaggregation, memory resources can be made more independent of computing resource expansion, improving GPU utilization and reducing data movement latency. The products released include:Bravera SC6 PCIe 6.0 SSD controller: Designed for AI inference storage scenarios, it helps cloud service providers migrate more KV Cache to high-performance SSDs, enhancing infrastructure efficiency. This product features an architecture compatible with multi-vendor NAND and is expected to begin sampling in the fourth quarter of 2026.Marvell Structera X memory expansion solution: Based on CXL technology, it supports rack-level memory expansion and resource pooling, helping data centers share and allocate memory resources more flexibly, reducing AI infrastructure costs.Marvell Photonic Fabric optical interconnect memory solution: Constructs a shared memory architecture across multiple racks using optical interconnect technology, supporting up to 32TB warm KV Cache offloading and helping AI inference clusters enhance throughput capacity.Marvell stated that the Photonic Fabric solution can achieve a 2 to 3 times increase in Token throughput under existing data center space and power consumption constraints, supporting larger scale models and longer context AI applications.Marvell executive Will Chu stated that AI infrastructure is transitioning from a single server architecture to a system where computing, memory, and connectivity operate in synergy, and in the future, memory needs to expand more independently to enhance resource utilization and Token efficiency.As the demand for AI Agents and large model inference continues to grow, memory capacity, bandwidth, and data transfer efficiency are becoming new focal points of competition in AI infrastructure, following computing power.

The global stablecoin payment infrastructure platform WasabiCard, in collaboration with WIDTH, held the CCO Leadership Edition compliance high-end exchange event in Hong Kong. WasabiCard CMO Jack Derong gathered with executives and experts from the fields of digital assets, payments, fintech, and regulatory compliance in Hong Kong to engage in in-depth discussions on topics such as the evolution of global regulations, the development of stablecoin payments, and corporate compliance practices.As Hong Kong's Web3 ecosystem continues to develop and the digital asset regulatory framework becomes increasingly refined, stablecoin payments are ushering in new development opportunities. With the continuous improvement of compliance requirements surrounding the HKMA stablecoin licensing system, Travel Rule, AML, KYT, and others, companies are increasingly focused on how to build payment infrastructure that combines global payment capabilities, risk management capabilities, and compliance capabilities to achieve the scalable implementation of stablecoin payments.WasabiCard believes that the future competition in stablecoin payments will hinge on whether companies possess the infrastructure capabilities to cover global compliance, risk management, and cross-border payment networks, rather than just payment efficiency.In the future, WasabiCard will continue to enhance global payment infrastructure based on the highest standards of compliance systems and risk management capabilities, helping clients confidently respond to the evolving global regulatory requirements and expand their international business more securely.

According to Bloomberg, Micron officially broke ground on its factory expansion project in Higashi-Hiroshima City, Hiroshima Prefecture, Japan on Saturday. The total investment for the project is approximately 1.5 trillion yen (about 9.3 billion USD), aimed at producing advanced memory chips for the AI era.The new production line will focus on key chips such as high-bandwidth memory (HBM), which are core components supporting the performance of AI processors like those from Nvidia. The related chips are expected to start shipping as early as the summer of 2028.The Japanese Ministry of Economy, Trade and Industry has provided up to 500 billion yen in subsidy support for the project to strengthen domestic semiconductor manufacturing capabilities and enhance supply chain resilience.This expansion project is seen as an important step for Japan to strengthen its semiconductor manufacturing layout in the global AI chip competition.

According to on-chain analyst Ai Yi's monitoring, the address that built a position of 3845 ETH at $2074 yesterday has added another 967.01 ETH. Since yesterday, a new round of accumulation has gathered 4812.03 ETH, with a total value of $10 million and an average acquisition price of $2078.91, currently showing an unrealized profit of $114,000.

Vitalik Buterin posted on the X platform that enhancing bandwidth is safer and more reliable than reducing latency. With the help of PeerDAS and ZKP, Ethereum's scalability can increase by thousands of times, and this does not conflict with decentralization. However, latency is constrained by physical laws and real-world environments, especially when supporting global home nodes, censorship resistance, and anonymity.He pointed out that reducing latency to a range of 2-4 seconds is a reasonable goal, while applications requiring higher real-time performance (such as AI-driven ones) will rely on L2 and "city chains" for localized scaling.

According to ChainCatcher news, the Bitcoin Layer2 network Stacks has completed an upgrade that expands network bandwidth by 500%. Core developers plan to continue with several other optimizations in 2025.ChainCatcher previously reported that the Nakamoto upgrade of the Bitcoin Layer2 network Stacks has been successfully launched. This upgrade will introduce "Fast Blocks" and a "Bitcoin Finality" mechanism to enhance transaction speed. After the Nakamoto upgrade, Stacks block production will be independent of Bitcoin block production, allowing user transactions to be confirmed in a matter of seconds.

ChainCatcher message, Titan Network breaks through technical and financial channels through innovative resource aggregation and business models. It addresses industry pain points in the DePIN field, such as difficulties in resource monetization, scaling, and poor commercial sustainability.Titan will launch the fourth round of the testnet on January 6, 2025, focusing entirely on the large-scale commercialization of idle bandwidth, aiming to achieve a monthly revenue of $300,000 and create a new paradigm for the commercialization of DePIN projects.The testnet is now open for resource access applications, and bandwidth resource holders and commercial clients can now join.

ChainCatcher news, Solana co-founder Toly releases the Solana roadmap, the document shows "increase bandwidth, reduce latency".

ChainCatcher message, Ethereum co-founder Vitalik Buterin posted on Farcaster that perhaps developers should realize that 32 ETH is a higher threshold for stakers than bandwidth requirements, and maybe they could slightly increase the bandwidth requirements in exchange for lowering the minimum staking deposit to, for example, 16 or 24 ETH. This would benefit both staking accessibility and scale. Then, once Ethereum is ready for the peerdas upgrade, the bandwidth requirements will decrease, and the minimum deposit limit could also be lowered to 1 ETH.

ChainCatcher news, according to official sources, the decentralized internet bandwidth sharing network Multisynq has announced the completion of a $2.2 million seed round financing. This round was led by Manifold, with participation from ARKN Ventures, PHD Capital, EnigmaFunge, AlphaCrypto, Republic Crypto, and a group of angel investors.It is reported that Multisynq enables individuals to sell excess internet bandwidth to developers, thereby monetizing it and achieving the goal of internet bandwidth sharing.