# SSD Related Articles

HTX News Center provides the latest articles and in-depth analysis on "SSD", covering market trends, project updates, tech developments, and regulatory policies in the crypto industry.

From Hot Storage to Cold Memory: Decentralized Storage in the AI Era's Storage Boom

"From Hot Storage to Cold Memory: Decentralized Storage in the Era of AI Storage Boom" This article explores the divergent market trajectories of AI-driven centralized storage and Web3's decentralized storage. It argues that while AI storage is experiencing a massive revaluation focused on "hot data efficiency" — maximizing computational throughput via technologies like HBM, enterprise SSDs, and sophisticated data pipelines — decentralized storage projects like Filecoin and Arweave are currently sidelined. Their core value proposition lies in "cold data trust," prioritizing data integrity, censorship resistance, and long-term archival over raw speed. The piece details the AI storage architecture, emphasizing its role as a "performance engine" critical for feeding GPUs, contrasted with decentralized storage's focus on serving as a permanent, verifiable ledger for humanity's collective memory. It analyzes the challenges decentralized storage faces, including product-market fit, enterprise readiness, and token economic misalignment, but concludes that its fundamental value in preserving provenance, public datasets, and civilizational archives positions it for potential long-term revaluation as issues of data sovereignty, AI auditability, and historical preservation become more acute. The current market rewards efficiency, but the pendulum may eventually swing back towards trust.

marsbit11h ago

From Hot Storage to Cold Memory: Decentralized Storage in the AI Era's Storage Boom

marsbit11h ago

A Year Consumes a Solid-State Drive: Codex Log Bug Slammed as 'Slopware'

OpenAI's flagship AI coding tool, Codex, was found to have a critical bug causing its feedback logging system to silently and rapidly wear out users' SSDs. A developer reported that Codex was writing approximately 640 TB of data per year to a local SQLite database (`logs_2.sqlite`) through a constant cycle of inserting and immediately deleting log entries, primarily at the verbose TRACE level. While the database file itself remained around 1 GB, the underlying write-amplification from SQLite's WAL mechanism meant the physical SSD endured the full write load. This was enough to exceed the typical 600 TBW endurance rating of a consumer SSD within a year. The root cause was a hardcoded default logging level (`Level::TRACE`) in the configuration, which overrode any user attempts to reduce logging via environment variables. Analysis showed that over 96% of the logged data—including noisy WebSocket packet dumps and repeated system file events—was useless debug information. The issue, which had at least nine related bug reports in the Codex repository, remained latent because it didn't visibly consume disk space, only silently accumulated write cycles. After the report gained traction on Hacker News, OpenAI merged fixes estimated to reduce writes by about 85%. However, even post-fix, the tool would still write an estimated 96 TB annually. The incident sparked broader criticism of "slopware" in AI-assisted development tools, highlighting a lack of resource budgeting for disk, CPU, and memory in always-on agent software, and a reliance on modern hardware to mask inefficient code. Competing tools like Claude Code were noted to have similar issues.

marsbit07/02 08:50

A Year Consumes a Solid-State Drive: Codex Log Bug Slammed as 'Slopware'

marsbit07/02 08:50

Research Report Analysis: Morgan Stanley Details SanDisk SNDK, The Truth About Cloud Data Center Pricing Power and AI Inference Benefits

Morgan Stanley raised its price target for SanDisk (SNDK) from $1100 to $1750 on June 22, maintaining an Overweight rating. The upgrade is driven by AI inference demand reshaping the NAND market, particularly for KV Cache and context window storage in cloud data centers. These cloud clients exhibit price inelasticity and sign long-term contracts, granting SanDisk significant pricing power. SanDisk's New Business Model (NBM) agreements, covering over one-third of FY27 bit shipments with 3-5 year terms and fixed price/price collar structures, are crucial. They are projected to sustain gross margins around 80% even at floor prices, providing a buffer against cyclical downturns. Morgan Stanley forecasts gross margins to surge from 30.3% in FY25 to 86.7% in FY27e. With NAND supply expected to remain tight into 2026/2027 and cloud/data centers becoming the largest end-market, SanDisk holds supply-side pricing power. The company targets 15-19% bit growth via technology transitions, not capacity expansion. Revenue is projected to grow ~6.6x from FY25 to FY27, with EPS rising from $2.74 to $14.73, driven by high-margin cloud business. Key upside catalysts include faster enterprise SSD adoption and edge AI growth. Downside risks involve slower industry growth, competitor capex increases, market share loss, and competition from Chinese players like YMTC. The investment thesis rests on AI-driven structural demand, NBM's margin protection, and sustained supply tightness. The $1750 target implies ~28x FY27e P/E.

marsbit06/23 12:04

Research Report Analysis: Morgan Stanley Details SanDisk SNDK, The Truth About Cloud Data Center Pricing Power and AI Inference Benefits

marsbit06/23 12:04

One Article to Understand the Profit Pools and Industry Landscape of the AI Storage Hierarchy

**Deciphering the Profit Pools and Industry Landscape of the AI Storage Hierarchy** AI storage architecture can be divided into six distinct layers based on proximity to computing units: 1) On-chip SRAM, 2) HBM, 3) Motherboard DRAM, 4) CXL pooling layer, 5) Enterprise SSD, and 6) NAS & Cloud Object Storage. In 2025, the total market for these layers (excluding embedded SRAM value) was approximately $229 billion, with DRAM constituting half, HBM 15%, and SSD 11%. The profit landscape is highly concentrated, with over 90% market share in the top three layers for key players. These profit pools are categorized into three types: 1) High-margin, oligopolistic silicon layers (HBM, embedded SRAM, QLC SSD), 2) High-margin, emerging interconnect layers (CXL), and 3) Scalable, recurring-revenue service layers (NAS, Cloud Object Storage). **Key Layers Analysis:** * **On-chip SRAM:** Profits accrue primarily to TSMC via advanced wafer sales for AI chips. * **HBM:** The largest AI-era profit pool, driven by AI accelerator demand. SK Hynix (57-62% share), Samsung, and Micron dominate. HBM boasts exceptionally high margins (e.g., SK Hynix's 72% operating margin in Q1 2026) and is projected to grow at a ~40% CAGR to $100 billion by 2028. * **Motherboard DRAM:** The largest market by revenue ($121.8B in 2025), controlled by Samsung, SK Hynix, and Micron. High profitability is sustained as capacity shifts to HBM. * **CXL Pooling Layer:** Enables rack-level memory sharing for AI workloads. The market is forecast to grow from $1.6B in 2024 to $23.7B by 2033. While memory giants lead, companies like Astera Labs (holding ~55% share in retimers/controllers) achieve very high margins (~76%). * **Enterprise SSD:** A major beneficiary of the AI inference era, especially QLC SSDs, with the market expected to reach $76B by 2030. Samsung, SK Hynix (including Solidigm), and Micron are key players. * **NAS & Cloud Object Storage:** The outermost data lake layer, growing steadily (CAGR ~16-17%). Profit derives from long-term data hosting, egress fees, and ecosystem lock-in, led by vendors like NetApp, Dell, and cloud providers (AWS, Azure, Google Cloud). **Summary:** Profitability correlates strongly with proximity to compute: layers like HBM and CXL components command the highest margins (60%+ and 76%+, respectively) despite smaller market sizes, while DRAM has the largest revenue base. The primary growth vectors are HBM (CAGR ~28%), Enterprise SSD (CAGR ~24%), and CXL pooling (CAGR ~37%). Barriers vary by layer, encompassing advanced manufacturing (HBM), IP/certification (CXL), and high switching costs (service layers).

marsbit05/14 04:03

One Article to Understand the Profit Pools and Industry Landscape of the AI Storage Hierarchy

marsbit05/14 04:03

Memory Card Prices Double in Four Months: How Long Will the Surge Last?

NAND flash memory prices have entered a rapid upward cycle, with consumer-grade storage products like microSD cards seeing significant retail price increases. For example, a SanDisk Extreme 128GB microSD card rose from $17 in October 2025 to nearly $40 by February 2026—a 130% surge in under four months. This price surge is driven by structural shifts in the NAND market, primarily due to soaring demand from AI data centers. These large-scale buyers are securing the majority of NAND wafer supply through long-term contracts, leaving limited inventory for the consumer market. According to TrendForce, NAND contract prices rose 55–60% in Q1 2026, with enterprise SSD prices climbing 53–58%. Retail prices rose even more sharply due to constrained supply in the distribution channel. Unlike the 2016–2017 price cycle caused by production transitions, the current spike is demand-led. AI data centers are consuming NAND capacity at an unprecedented rate, with 2026 demand growth estimated at 20–22% against supply growth of only 15–17%. Manufacturers are prioritizing high-margin enterprise products over consumer-grade storage, further tightening retail availability. New production capacity from major suppliers like Samsung, Micron, and Kioxia is not expected until late 2027 or 2028. Until then, consumer storage prices are likely to remain high, with no significant price relief anticipated in the near term.

marsbit04/16 03:13

Memory Card Prices Double in Four Months: How Long Will the Surge Last?

marsbit04/16 03:13

活动图片