agi

Goldman Sachs announced the acquisition of NEOS Investments, which is an ETF issuer managing assets totaling $30 billion, including the Bitcoin High Income ETF (BTCI), which manages assets of $1.1 billion.

ChianCatcher news, Ethena announced a partnership with digital asset prime broker FalconX, becoming its institutional lending partner. As part of the collaboration, Ethena will include stablecoin lending arrangements as part of its institutional lending configuration, investing in FalconX through over-collateralization.FalconX can use this to expand its balance sheet, while Ethena benefits from FalconX's expertise in loan origination and secured lending, thus obtaining risk-adjusted terms superior to other channels.

According to Taiwanese media reports, Intel's next-generation advanced packaging technology EMIB-T is facing severe yield challenges. ABF substrate supplier Unimicron stated during its earnings call on July 29 that the technology is "not yet mature," and the yield target for three EMIB-T substrate suppliers (Unimicron, Ibiden, and Shinkawa) during the initial mass production phase at the end of 2027 is only 50%.EMIB-T is Intel's advanced packaging solution that competes with TSMC's CoWoS-L, differing in that it embeds the silicon bridge directly into the substrate rather than using a silicon interposer, making substrate suppliers a key factor in determining success or failure. Google's ninth-generation TPU has confirmed that it will adopt EMIB-T packaging when it goes into mass production in 2028, with MediaTek responsible for chip co-design. If the chip successfully goes into mass production, its shipment volume is expected to challenge NVIDIA. Unimicron stated that the customer (Intel) has provided a profit guarantee mechanism, ensuring that suppliers can remain profitable even if yields are poor, and that profit margins will exceed the company's average level after achieving the 50% yield target. Currently, there is still considerable uncertainty about whether EMIB-T can achieve its mass production target by the end of 2027.

According to the Electronic Times, TSMC plans to further expand the outsourcing scale of the CoW (Chip-on-Wafer) segment in its CoWoS advanced packaging, entrusting more packaging orders to ASE and other OSAT manufacturers. CoWoS is TSMC's 2.5D packaging technology that connects AI processors such as GPUs with HBM through an intermediary layer, where CoW is the step of combining the chip with the intermediary layer, and WoS is the step of packaging the substrate. Previously, TSMC mainly outsourced the WoS segment, with only a small portion of CoW being entrusted.This move aims to alleviate the packaging capacity bottleneck caused by the continuous surge in demand for AI chips. TSMC holds about 90% of the global AI chip manufacturing market, but as the demand for self-developed chips from manufacturers like Nvidia and AI data centers skyrockets, capacity has become difficult to meet. Industry insiders expect OSATs to significantly increase equipment investment, especially in post-process areas such as wafer cutting and bonding. Several laser processing and bonding equipment suppliers in South Korea have been confirmed to be negotiating with OSATs regarding the supply of CoW equipment.

The market research agency The Kobeissi Letter indicates that investors are accelerating the allocation of tokenized U.S. Treasuries, driving the continuous expansion of the on-chain U.S. Treasury market. Data shows that the total market value of on-chain U.S. Treasury funds has risen to a historical high of $16.2 billion, an increase of approximately 77% since the beginning of the year.Kobeissi points out that market growth is primarily driven by demand for on-chain yields. An increasing number of users are using tokenized U.S. Treasuries as collateral to borrow stablecoins and deploy funds into DeFi strategies to enhance capital efficiency. Some on-chain lending platforms support looping leverage strategies, allowing users to repeatedly collateralize tokenized Treasuries, borrow stablecoins, and reinvest in the market, with some strategies offering annualized returns exceeding 10%.Kobeissi believes that as traditional financial assets continue to go on-chain, tokenized U.S. Treasuries are becoming an important infrastructure for on-chain finance and may become a significant component of future on-chain capital markets.

According to SolanaFloor, Solana's "Resource and Inclusion Fee" proposal SGP-0003 has entered the support stage. If it receives 15% active staking support, the proposal will proceed to a full governance vote.If implemented, the daily SOL burn amount is expected to increase from the current 650 coins (approximately $51,000) to a maximum of 9,000 coins (approximately $702,000).

Dongfang Harbor Chairman Dan Bin expressed his views this morning, stating that in July, the chip sector experienced a sharp decline and massive leveraged liquidations, but in the grand AI cycle, such a level of correction is both a necessary path and a sign of market health.The market has yet to fully understand the infinite demand potential of AI as an "intelligent" product. Concerns about capital expenditures from giants have echoed the early story of Amazon AWS, but the opportunities in AI are far greater. Funds are flowing back from low-quality tech stocks to high-quality targets, confirming the judgment of a "return of the king" by the end of 2026. Storage chips still face cyclical risks and high volatility, so it is advisable to wait for technical repairs, with a more optimistic view on fundamentally solid companies like Nvidia, Broadcom, and TSMC, as funds will flow more towards quality application layers.On the other hand, the business of hyperscale cloud providers is accelerating growth, with a large backlog of orders and an increasing growth rate, indicating that the previous punishment of capital expenditures was a misjudgment, and these investments will translate into certain future revenues. Looking ahead to August, the Nasdaq's rebound is expected to continue until Nvidia's earnings report, and the tech sector's rolling adjustment is nearing its end, with funds accelerating back into high-quality tech stocks.In terms of specific sector operations, it is currently not advisable to blindly chase high prices in the chip and storage (memory) sectors. However, investors can adopt a short-term band trading strategy of buying on dips until the storage sector completely emerges from its bottoming pattern on a technical level.

According to X user Rohan Paul, when the OpenAI model breached Hugging Face, Anthropic's Claude 5 refused to assist in the investigation, and Hugging Face ultimately turned to Nvidia's quantized version of GLM 5.2 for emergency response.The user commented that banning open-source models would weaken the defense's capabilities, and this incident confirms the practical value of open-source models in security response.

According to The Information, TSMC (TSM.N) will develop AI chip packaging technology.In the past, packaging has been regarded as a relatively conventional process in semiconductor manufacturing, but with the surge in AI demand, chip design companies need to integrate more processors and high-bandwidth memory into increasingly large and complex packages. Advanced packaging has become one of the industry's most critical bottlenecks. Intel has developed a packaging technology called Embedded Multi-Die Interconnect Bridge (EMIB).This technology uses small silicon bridges to establish high-speed connections between processors and memory, enabling support for larger-scale multi-chip designs, thus creating more powerful AI chips and helping chip designers achieve supply chain diversification, which has attracted the attention of potential customers such as NVIDIA (NVDA.O).Insiders say that within TSMC, engineers refer to this advanced packaging project as "EMIB-like" and have discussed this solution with some customers.

According to Jinshi News, the European Union has launched a tender today, planning to establish up to 7 artificial intelligence super factories in Europe, which is an important initiative to promote technological autonomy in Europe. The plan is led by the industry and has received up to 10 billion euros in funding support, expected to leverage at least 20 billion euros in private investment.AI super factories will provide infrastructure for various enterprises and institutions, integrating advanced AI processors, software, and cloud technology, further strengthening Europe's leadership position in the technology sector.

Arm released its financial report for the first quarter of fiscal year 2027, with revenue of $1.29 billion, a year-on-year increase of 22%, exceeding the market expectation of $1.26 billion; adjusted earnings per share of $0.45, a year-on-year increase of 29%, surpassing the expected $0.40; adjusted net profit of $480 million, a year-on-year increase of 28%. Licensing revenue was $574 million, a year-on-year increase of 23%; royalty revenue was $715 million, a year-on-year increase of 22%, with data center royalty revenue more than doubling year-on-year.The AI and data center business is showing strong momentum. Since the launch of the Arm AGI CPU in March, customer demand has grown from an initial expected order of about $1 billion to over $2 billion. Cumulative shipments of Arm Neoverse data center processors have exceeded 1.5 billion cores, with the latest 500 million completed in just 9 months. NVIDIA Vera has fully entered production, and AWS has reached a multi-year agreement with Meta to deploy tens of millions of Graviton5 cores. Qualcomm announced plans to launch the Arm-based data center CPU Dragonfly C1000.Revenue guidance for the second quarter is approximately $1.38 billion, with adjusted earnings per share of about $0.47, both above market average expectations but below some optimistic forecasts. The company stated that the growth of smart mobile royalty revenue will slow, and after-hours stock prices fell by about 7% at one point.

According to TNW, the Cyberspace Administration of China has released a draft of regulations for combating cyberbullying and is seeking public feedback. The draft requires platforms to establish classifiers for detecting and categorizing online violence content, provide users with a "one-click protection" feature, and implement blacklist management and regular reporting mechanisms for violating accounts.Notably, the draft explicitly includes AI-generated deepfake images and synthetic harassment texts within the scope of regulation. The Q&A platform Zhihu has already responded by restricting the use of anonymous accounts.

According to The Block, Magic Labs has agreed to sell its embedded wallet business to cryptocurrency exchange Kraken's parent company Payward as part of a broader company restructuring, with Magic Labs simultaneously rebranding to Newton Labs. This transaction is in the form of an asset sale, and after the deal is completed, Magic Labs and Payward will continue to operate independently, with wallet customers being taken over by Payward Services.Since its establishment in 2018, Magic Labs has created over 60 million wallets, supporting more than 200,000 developers. Newton Labs will focus on developing the Newton Protocol, which is an authorization layer for on-chain finance that executes compliance, security, and risk strategies before transactions are settled on-chain. The protocol entered the mainnet testing phase in June 2026. The first product based on the Newton Protocol, VaultKit, is a set of composable strategy tools for institutional-grade vaults.This acquisition is another transaction in Payward's ongoing expansion of its financial infrastructure platform. Previously, the company completed the acquisition of the CFTC-licensed derivatives exchange Bitnomial for up to $550 million in cash and stock, and acquired the Hong Kong stablecoin payment company Reap Technologies for $600 million. In 2025, it also acquired the retail futures platform NinjaTrader for $1.5 billion.

According to monitoring by Catcher Predict, in the predicted market Polymarket for the event "Counter-Strike: magic vs 3DMAX - BLAST Bounty Qualifier," the "magic" option in the sub-market "Map 2 Winner" has experienced a dramatic fluctuation in winning probability, dropping from 67.5% an hour ago to the current 38.5% (a fluctuation of 29%). Please note the impact of related breaking news.

Domestic DRAM storage IDM leader Changxin Technology will officially list on the Sci-Tech Innovation Board on July 27, with an issuance valuation of approximately 580 billion yuan and a total fundraising amount of 29.5 billion yuan, of which about 22.066 billion yuan is for equipment purchase and installation costs. Industry insiders believe that 2026-2027 is the golden window period for the introduction of domestic equipment.In terms of the supply chain, on the equipment side, North Huachuang has shipped a large volume of various categories such as etching and thin film deposition, Zhongwei Company continues to validate and mass-produce etching and thin film equipment, Huahai Qingshi CMP equipment has entered the core supply chain, and Jingce Electronics and Jingzhida testing equipment are benefiting simultaneously. On the materials side, Yake Technology precursor products cover advanced process needs, Guanggang Gas provides special gases and bulk gases, Jinhong Gas has a gas production order cycle of 15 years, Tongcheng New Materials ArF/KrF photoresists have entered mass production, Jingrui Electric Materials has a market share of over 40% in high-purity hydrogen peroxide, but high-end photoresists are still in the validation stage, and Debang Technology is in the sample testing stage. On the packaging and module side, Huatian Technology is a packaging supplier, Jiangbolong and Demingli have reached procurement agreements, while Maijie Technology is still in the introduction phase. On the distribution side, Shangluo Electronics has been authorized, and there are expectations for supply strategy adjustments after listing. Several listed companies hold shares indirectly through industrial funds, with Shangfeng Cement holding approximately 0.1517% of the equity before issuance.

According to Reuters, Amazon conducted layoffs in its Artificial General Intelligence (AGI) department on July 22, marking a new round of smaller-scale adjustments following the company's reduction of 16,000 employees in January of this year. A company spokesperson stated in response to Reuters' inquiry that AGI remains one of the company's most important directions, but it had to make the "difficult decision" to eliminate certain positions to focus on customer priority projects.The layoffs involve the teams overseen by AGI Data Services Vice President Adeeb Shanaa and AGI Information Vice President Vishal Sharma, with specific details on the scale of the cuts not yet disclosed. Previously, Amazon AGI head Rohit Prasad left at the end of last year, and AGI Lab head David Luan departed in February of this year. AGI-related operations were integrated under Senior Vice President Peter DeSantis in December of last year, along with departments such as chip development and quantum computing.

Last month, comments from Liang Wenfeng during the DeepSeek investor meeting have recently circulated. Liang Wenfeng clearly stated that DeepSeek's goal is AGI, rather than becoming the next super app. They will not engage in 3D, video generation, or world modeling, nor will they close source code, stating, "Restraint is a strategy used to increase the probability of achieving AGI." He believes that products are merely "by-products" on the path to AGI, with the current focus on Coding Agent and continuous learning capabilities. The next phase aims to achieve autonomous model iteration, with the endpoint pointing towards embodied intelligence.In terms of commercialization, Liang Wenfeng stated that they only seek to earn "reasonable profits" and do not set profit maximization as their pricing goal. DeepSeek is "still far from a complete shift to commercialization." He assesses that the gap between AI in China and the U.S. mainly lies in resources rather than talent, stating, "Using a fraction of the computing power can narrow the gap to 6 months, 3 months." Regarding industry competition, he believes that open source is "the sweet point for companies of this scale," while also predicting that domestic large models will ultimately converge to "two large companies and two small companies." Liang Wenfeng also mentioned that maintaining team stability is the only non-negotiable core principle, and the company is driven by vision, with "a group of ordinary people achieving extraordinary things."