Core Insights - The rapid growth of data in the AI era is highlighted, with global data expected to increase by 36% annually, surpassing 1 YB by 2030, marking the entry into the YB era [3][4] - In China, the annual data output reaches 41.06 ZB, but only 5.1% is effectively stored, indicating a significant opportunity for improvement in data utilization [4] - The establishment of data aggregation platforms is being promoted by national policies to transition data from fragmentation to centralization, enhancing the value of data elements [4] Group 1: Data Infrastructure Development - The concept of "storage power centers" is introduced as a new type of data infrastructure, focusing on comprehensive management of data through aggregation, governance, security, and application [4][5] - Huawei's Omni-Dataverse global file system facilitates efficient data flow across data centers and regions, ensuring data visibility and manageability [5] - The ModelEngine toolchain enhances data preprocessing efficiency and reduces model costs, accelerating AI application deployment [5] Group 2: Data Security and Governance - Huawei's storage centers implement a trusted data space with over 30 data control strategies and full-process logging to ensure secure and controllable data circulation [5] - The integration of end-to-end data enterprises within storage centers promotes a positive cycle from data collection to governance and application [5] Group 3: Strategic Goals for Data Utilization - The goal for China is to transition from a data powerhouse to a data stronghold by increasing the application of advanced flash storage across industries and encouraging innovation in new storage mediums [5] - There is a call for enhanced deployment of AI storage facilities and improved data disaster recovery coverage [5] - The official launch of the provincial advanced storage power center marks a new phase in data resource integration and value release, laying a foundation for market-oriented data element allocation [5]

Core Insights - The research team at Tianjin University has developed a new DNA storage system called HELIX, specifically designed for storing biomedical data, achieving the storage and recovery of 60MB of spatiotemporal omics images [1][2] - Traditional storage methods are becoming inadequate in the era of big data, leading to the emergence of DNA information storage technology, which is seen as a potential solution for large-scale data storage [1] - DNA can store hundreds of exabytes of data per gram and can preserve information for thousands of years without power, making it particularly promising for biomedical data applications [1] System Features - HELIX consists of three core modules: image compression, error correction coding, and image restoration, enhancing the system's fault tolerance and recovery capabilities [1][2] - The team optimized existing compression algorithms to address potential base errors during DNA storage, significantly improving the system's robustness [2] - Deep learning techniques were incorporated to enhance the success rate of image decoding and information recovery during the image restoration process [1][2] Experimental Results - In wet lab experiments, the team successfully encoded two 60MB spatiotemporal omics images into 130,000 DNA sequences, each containing 183 bases, and recovered the image data using DNA synthesis and sequencing technology [2] - The HELIX system demonstrated strong robustness, requiring only about 5.8 times the sequencing depth to recover the majority of the image information [2] - This achievement marks a significant step towards the practical application of DNA information storage technology, showcasing superior storage efficiency and reliability [2]

Company News - Super Micro Computer (SMCI.O) experienced a pre-market drop of nearly 17% after reporting preliminary Q3 revenue of $4.5 to $4.6 billion, significantly below analyst expectations of $5.35 billion [1] - Microsoft (MSFT.O) plans to actively contest any government orders that may suspend or halt its cloud business in Europe, while expanding its data center operations in 16 European countries [2] - Alibaba (BABA.N) is set to announce its Q1 2025 results on May 15, 2025 [2] - Stellantis (STLA.N) reported Q1 revenue of €35.8 billion, a 14% year-over-year decline, falling short of analyst predictions of €35.4 billion, and has suspended its 2025 financial guidance due to tariff-related uncertainties [2] - Visa (V.N) reported Q2 revenue of $9.6 billion for fiscal year 2025, exceeding market expectations of $9.55 billion and up from $8.78 billion in the same period last year [2] - Vale (VALE.N) CEO stated that the company reversed the decline in iron ore production in April and is confident in meeting its annual production guidance [2] Economic Indicators - The initial estimate for the U.S. Q1 real GDP annualized rate recorded at -0.3%, marking the lowest level since Q2 2022 [3] - Kalshi, a financial trading and prediction market platform, currently estimates a 74% probability of a U.S. economic recession this year [3] - The ADP employment report for April showed an increase of 62,000 jobs, the smallest gain since July 2024, and significantly below expectations [3] - Starbucks (SBUX.O) reported Q2 earnings of $384 million, halving compared to the same period last year, with same-store sales declining for the fifth consecutive quarter [3] - Taiwan Semiconductor Manufacturing Company (TSM.N) has begun construction on its third wafer fab in Arizona, increasing its expansion efforts in the U.S. [3] - Western Digital (WDC.O) saw a pre-market increase of 11.5% due to its Q4 revenue and profit guidance exceeding expectations [3]