Recently, the "First Domestic Storage Stock" ChangXin Storage officially landed on the GEM and set off a frenzy with a stunning 500% surge. Despite the recent overall pullback in the storage sector, AI storage is still being frenziedly revalued by capital in the current tech narrative wave. Meanwhile, decentralized storage in the Web3 field has plunged into long-term silence and loss. Why is there such a stark contrast in market performance under the shared name of "storage"? The fundamental answer lies in the complete divergence of underlying value functions.
The revaluation of storage in the AI era is fundamentally a carnival about “hot data efficiency”, which serves the ultimate maximization of computing power utilization and commercial realization; whereas what decentralized storage adheres to is the value proposition of “cold data trust”, defending data fairness, resistance to censorship, and the long-term memory of human civilization. The former is an efficiency system for hot data, while the latter is a trust system for cold data. The current capital market undoubtedly firmly stands on the side of "efficiency," but human civilization still ultimately needs an immutable memory foundation. The long-term value of trustworthy cold storage has never disappeared; it is only dormant on the dark side of the cycle, waiting to be revalued by the era.
Why Storage Has Once Again Become the Focus of the AI Industry Chain
In the traditional IT era, storage was a "capacity business." Enterprise CIOs focused on unit capacity cost, disk reliability, disaster recovery plans, archiving strategies, and a 3–5-year equipment refresh cycle. Storage was seen as an accessory purchased alongside servers.
This round of storage heat is not a traditional cycle recovery but a repricing by AI of data flow capabilities. In the era of large models, storage logic has transformed from "capacity-first" to "efficiency-first", relentlessly pursuing extreme metrics such as GPU saturation rate, Checkpoint writes, and RAG ultra-low latency. This marks the transformation of storage value from the "final resting place of data" to the "high-speed channel for data entry into computation."
The evolution of AI infrastructure bottlenecks is essentially a battle to fill the gaps in a "bucket effect." The real utilization rate of computing power is not a linear stacking of single assets but a stringent multiplier effect: Real Utilization Rate of Computing Power = GPU × HBM × DRAM × SSD × Network × File System. A bottleneck in any part will lead to a collapse of the overall computing power utilization. In the AI era, storage has transitioned for the first time from a "cost center" to an "efficiency engine." This is the fundamental logic behind the repricing of storage.
AI Storage Architecture Landscape: From HBM Bandwidth Organ to Data Lake Foundation
AI storage is by no means a simple hardware stack-up, but a tightly coupled, hierarchically scheduled complex system. In this system, industrial value and capital focus are highly concentrated on HBM, enterprise SSDs, SSD controllers, NVMe/CXL protocols, and high-performance storage systems. To clearly break down its value flow, we divide the AI storage architecture from top to bottom into four core levels:
· Compute Near-Memory Layer (Bandwidth Core): Dominated by HBM, supplemented by DRAM and CXL memory pooling technology. This layer is directly integrated into the GPU/CPU package or bus, aiming to break the "memory wall" and is the first gateway to determine whether computing power can be fully unleashed.
· High-Speed Persistent Storage Layer (IO Hub): The core logic consists of enterprise SSD = NAND Flash + SSD controller + NVMe/PCIe data path. This layer handles high-frequency checkpoint writes, massive training set loading, and RAG hot data caching, making it the most explicit persistent storage increment in AI data centers.
· Low-Cost High-Capacity Storage Layer (Capacity Foundation): Comprised of HDDs, cold storage, and data lake archiving systems. Faced with exponentially expanding multi-modal raw data, historical logs, and compliance backups, this layer still provides an irreplaceable TCO (Total Cost of Ownership) advantage.
· AI Storage Systems and Data Software (Scheduling Brain): Including high-performance parallel file systems, distributed object storage, vector databases, and RAG data governance layer. What AI truly consumes is not raw hardware but data availability that is efficiently organized, indexed, and permissioned by the software stack.
· As an ecological extension, decentralized storage is not directly involved in the millisecond-level race of AI hot data but instead anchors public dataset attestation, AI training data provenance, and long-term cold memory archiving, establishing its unique ecological position as the "trusted cold layer."
HBM: The "Bandwidth Organ" Closest to Computational Power in the AI Storage Chain
High Bandwidth Memory (HBM) is not traditional storage but a high-bandwidth memory layer near the GPU. Its core mission is not to store data but to continuously "feed" data to computation with extremely high bandwidth. HBM is the AI storage chain's segment closest to computational power, with the highest determinism, directly determining whether the GPU can be "well-fed," making it the current most critical supply chain bottleneck.
The HBM core architecture is based on "3D DRAM Stacking + 2.5D Advanced Packaging": By leveraging TSV vertical stacking and CoWoS heterogeneous integration, the storage-compute distance is dramatically reduced, enabling a leap in bandwidth at the differential level. Its industry barrier is not just DRAM design, but the system engineering of DRAM process, TSV, ultrathin stacking, packaging, thermal management, testing, and customer certification. Any yield defect in any of these aspects will result in the scrapping of the entire HBM Stack.
Currently, globally, only SK Hynix, Samsung, Micron—the three giants—can achieve stable mass production, establishing a triple moat of top-notch DRAM process, packaging capabilities, and NVIDIA/AMD customer certification.
DRAM and CXL: System Memory Foundation and Memory Pooling Engine
HBM addresses the GPU's near-end bandwidth limit, DRAM solidifies the server system's memory foundation, while CXL attempts to break the physical boundaries, restructuring the organization of data center memory resources.
· DRAM: Mainly responsible for CPU-side cache, data preprocessing, intermediate state staging, and system operation, it is the most fundamental system memory layer in servers. The global DRAM market is highly concentrated among the three giants SK Hynix, Samsung, and Micron; ChangXin Memory Technology (CXMT) is the core variable for China's domestically produced DRAM substitution.
· CXL (Compute Express Link): It is a new-generation cache-coherent interconnect protocol for data centers aimed at breaking through the limitations of traditional DIMM slots, local memory capacity, and the isolated nature of server memory resources, promoting the evolution of memory architecture towards expansion, pooling, and sharing. Currently, CXL is still in the early stage of transitioning from platform support to large-scale deployment, with high long-term architectural value; key companies include Astera Labs and LightBits Labs.
Enterprise SSD: Data Hub Built by NAND, Controller, and NVMe
Enterprise SSD is the core high-throughput persistent increment of AI data centers, providing extremely high throughput, ultra-low latency, and stable QoS, continuously "feeding" data to the GPU, spanning the entire lifecycle from training data loading, checkpoint writes, RAG retrieval, inference cache, to log replay.
In the AI storage architecture, SSD is not a standalone hardware but a highly coupled system that can be distilled into an industry formula: Enterprise SSD = NAND Flash + SSD Controller + NVMe/PCIe Data Path. The three layers represent independent links in the industry chain:
· NAND Flash (Raw Material Layer): Determines storage density and unit cost, where the controller oversees performance optimization and lifespan management. Companies represented include Samsung, SK hynix (Solidigm), Micron, Kioxia, Western Digital, and YMTC.
· SSD Controller (Performance Empowerment Layer): Determines performance optimization, data error correction, QoS stability, and wear leveling. Companies represented include Phison (Fison), Silicon Motion, Marvell, and Maxio (Silicon Integrated Systems).
· NVMe/PCIe (Data Path Layer): Determines the data transfer efficiency from storage to computation. In combination with GPUDirect Storage technology, it reduces CPU memory bounce buffer and CPU involvement, significantly alleviating I/O bottlenecks. Companies represented include Broadcom, Marvell, and Astera Labs.
HDD / Cold Storage / Archive: Low-Cost Foundation of AI Data Lake
AI will not eliminate HDD. With the multi-modal large models' demand for video and image data, as well as the exponential expansion of enterprise compliance logs and historical datasets, the need for low-cost cold data storage is synchronously surging. In the AI storage architecture, SSD and HDD collaborate in a tiered manner based on business value: SSD handles hot data and high throughput, while HDD addresses low cost and long-term retention. Companies represented include Seagate, Western Digital, and Toshiba.
AI Storage Software Stack: The Scheduling Hub of Data Availability
AI never truly consumes bare disks but rather "data services" that have been meticulously organized by the software stack. This architecture transforms the underlying hardware into knowledge assets that AI can directly invoke, divided into four layers:
· High-Performance Storage System (Feeding System): Core-focused on concurrent throughput and low latency, it resolves the "data starvation" issue of GPU clusters through a parallel file system, ensuring rapid flow between training and inference. Companies represented include VAST Data, WEKA, and Pure Storage.
· Object Storage (Raw Data Lake): Built on Object, Key, and Metadata management, it stores massive amounts of unstructured data. It prioritizes low cost and cloud-native features over ultra-low latency, establishing a capacity foundation. Representative company: AWS S3
· Vector Database (Semantic Index Layer): The vector database is responsible for storing, indexing, and retrieving vectors generated by embedding models, enabling AI to accurately locate relevant content from vast knowledge. Representative companies: Pinecone, Milvus
· RAG Data Layer (Knowledge Retrieval Layer): Going beyond single-point retrieval, it covers data slicing, cleansing, permission control, and reference lineage to ensure enterprise data can be securely, accurately, and traceably retrieved by large models. Representative company: Databricks
From AI Hot Storage to Decentralized Cold Memory: Maximizing Efficiency vs Maximizing Trust
AI storage is an efficiency-driven system, with its value function focusing on maximizing computational output. HBM bandwidth determines whether GPUs can be fully utilized, SSD throughput determines the efficiency of dataset and checkpoint read/write operations, and low latency is crucial for real-time RAG and inference experiences. These metrics ultimately converge into GPU utilization and per-token cost, directly impacting the business profitability of AI applications. The ultimate goal of AI storage is not preservation but acceleration, serving productivity.
On the other hand, the value function of decentralized storage is fundamentally different. It questions whether data will still exist ten years later, whether it has been tampered with, and whether it can resist single-point audits. Through cryptographic proofs and a distributed network, it creates an open-access and permanently stored public data foundation. Its ultimate goal is to uphold the absolute truth and sovereignty of data, serving fairness, censorship resistance, and collective memory.
AI storage is the "hot storage" that provides fuel for future productivity, while decentralized storage is the "cold memory" that preserves an immutable historical record of humanity. The former serves efficiency, pursuing ultimate speed; the latter serves trust, safeguarding silent memories. The former determines how fast models run, the latter determines whether memories will be erased. Currently, market mechanisms reward efficiency in productivity, placing AI storage at the forefront, while decentralized storage seems to be experiencing a collapse in valuation and a narrative bloodletting silence.
The Vision and Reality of Decentralized Storage
There are many decentralized storage projects, but based on industry awareness and ecosystem accumulation, the core representatives have always been Filecoin and Arweave. Although both belong to "decentralized storage," their underlying architectural philosophies are almost two completely different paths—where the former approaches AWS's elasticity through a market-driven contract, and the latter approaches the eternity of a library through a one-time social contract.
· Filecoin: It has built the most complete verifiable economic system through PoRep and PoSt. It should not continue to compete with AWS on consumer-grade cloud storage, but should pivot towards AI data provenance, public dataset hosting, and compliance archiving to provide verifiable chains for model auditing and copyright claims. The inevitable path is to encapsulate it into an S3-compatible API and support fiat payments, upgrading from a "cheap storage market" to a "verifiable computing infrastructure."
· Arweave: With the narrative of "pay once, store permanently," it incentivizes miners to store and quickly access as much historical data as possible, especially scarce data, through the Blockweave and SPoRA mechanisms. Its ideal position is the foundation of human public memory—preserving human rights records, evidence of war crimes, cultural heritage, archiving legal and financial history, providing AI agents with permanent access to long-term memory. Arweave's value lies not in speed but in its capacity to carry civilization's memory across epochs.
The dilemma faced by decentralized storage projects such as Filecoin and Arweave lies not in a flawed value proposition but in the long-term misalignment of productization, retrieval experience, real needs, and token incentives. This reveals a significant gap from geek ideals to mainstream business applications:
· Mismatched Supply-Demand Incentives: Early networks represented by Filecoin rapidly expanded through tokens but failed to build a strong enough demand-side payment, resulting in significant capacity but insufficient utilization and payment conversion. It rewards "I can store" rather than "I need to store."
· Lack of Enterprise-Grade Service Capability: The barrier of entry for AWS is not the hard drive but the "data operating system" consisting of API, SLA, permission management, compliance audits, and technical support. Enterprises purchase "peace of mind," not experimental infrastructure that requires handling keys and node selection.
· Retrieval Experience Shortcomings: "Storing in" does not equate to "reliably and with low latency retrieving." With decentralized nodes, complex topologies, and a lack of unified SLA, it is challenging to accommodate AI hot data workflows and is more suitable for trusted cold archiving and data provenance.
· Privacy Compliance Challenge: Enterprise private data cannot be easily written to a public immutable network; the right to be forgotten and immutability inherently conflict. Decentralized storage is more suitable for public data and long-term archives, rather than indiscriminately handling core private data.
· Tokenomics Hype Cycle: The bull market financialization masks insufficient demand, while the bear market sees miner ROI decline, exposing commercialization weaknesses. Tokens can bootstrap supply but cannot automatically generate demand and sustainable revenue.
On the other hand, other decentralized storage projects tend to focus on specific ecosystems or niche tracks: Storj/Sia, with cross-cycle industry expertise, have weaker Web3 narrative influence compared to Filecoin/Arweave; BNB Greenfield/Walrus are tied to specific public chain ecosystems with BNB or SUI; Celestia/EigenDA belong to the Data Availability (DA) layer, serving Rollup transaction confirmations rather than long-term archiving; 0G and other AI/DA hybrid narrative projects attempt to integrate storage, data availability, computation, and AI agent settlement into an AI-native modular infrastructure, but their actual demand, developer adoption, and commercialization loop are yet to be validated.
The Future Opportunities of Decentralized Storage: Efficiency and Trustworthy Long-term Pendulum
During the technology dividend explosion, capital frenzily pursued efficiency, with assets such as GPUs and HBM commanding a high premium, pushing decentralized storage advocating for "trustworthiness and fairness" to the sidelines. However, the pendulum of history will not stay forever at the efficiency end. Events such as super-platform arbitrary bans and content deletion, AI copyright litigations driving data provenance demands, geopolitical conflicts sparking data sovereignty disputes, data monopolies leading to public records disappearance, and regulatory audit pressures on model training data compliance could all brew a reevaluation of the pricing of "trustworthy storage." The future opportunities of decentralized storage still have the potential to demonstrate unique value in the following directions:
· AI Data Provenance: Building "data lineage proof" through cryptographic evidence to address regulatory and audit pressures.
· Public Datasets and Cultural Archives: Anchoring curated archives and cultural heritage to construct irreplaceable, undeletable memories.
· Trusted Archiving and Compliance Notarization: Achieving trustworthy self-certification through Hash attestation, providing high-level digital notarization.
· ZK/TEE/DID Technology Integration: Resolving privacy tensions, upgrading from a single "storage protocol" to a "trusted data infrastructure".
· Stealth Product Roadmap: Providing S3-compatible API and fiat billing, allowing users to directly purchase "trusted archive" services.
AI storage and decentralized storage, one pursuing ultimate efficiency to provide fuel for our journey to the future; the other defending silent memory, preserving our right to look back. The current market overwhelmingly rewards efficiency, hence decentralized storage appears silent and even collapsing; but as the AI era further magnifies data monopolies, copyright disputes, and the fragility of historical memory, decentralized storage may undergo a value reassessment in the form of "trusted cold storage". Memories that cannot be easily erased by platforms, companies, or any single power may transition from idealistic romanticism, from fringe beliefs, into essential infrastructure.
Welcome to join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Official Twitter Account: https://twitter.com/BlockBeatsAsia
