Shi Jing ping

As blockchain and decentralized systems continue to develop and expand, the importance of data storage has become stronger than ever before. For many years, centralized data storage systems like AWS S3, Google Drive, and Dropbox dominated the marketplace, but there has been an explosive increase in the availability of decentralized solutions. Of the many decentralized solutions available, one of the more advanced is called Walrus, a decentralized Filesystem that is built on top of the Sui Blockchain and the native token WAL fuels network rewards and data storage.
Walrus is a new and innovative Platform not simply a file system: it provides developers and users with a low-cost, robust, and programmable storage layer for their large data blobs (videos, datasets, NFTs, and models).
📌 What Sets Walrus Apart?
Rather than keeping entire data files located on one server or layer directly to the blockchain, Walrus breaks the data files into "pieces" with the use of sophisticated encodings, distributes such pieces throughout various decentralized nodes, and maintains blockchain stored proofs of those pieces.
This model has advantages such as:
Cost Efficiency: By eliminating unnecessary on-chain storage expenses.
Fault Tolerance: The ability to retrieve data if a significant number of decentralized nodes go offline.
Verifiable Storage: Available proofs to verify storage through the blockchain are accessible by every user.
📍 Storing Data on Walrus Information Store — Sequential Steps
Walrus Information Store consists of the following phases:
1. File Uploader + File Encoding
Upon uploading your File (this is the name of the uploaded file), Walrus Information Store does not necessarily keep your file on a One to One basis. Instead, it creates multiple files from one file stored in Walrus Information Store by cutting them into Multiple File Size Chunks using a Special Encoded Process called Red Stuff. Red Stuff is an Enhanced Version of Erasure Codes.
The Multiple File Size Chunks or Slivers) are distributed onto Multiple Independent Storage Nodes, meaning each Sliver will live independently across several different Storage Nodes.
To ensure that even if all of the Slivers are lost or destroyed, the Original File can still be reconstructed."} This is achieved by creating Redundant Parity Files that will be used to reconstruct the Original File.
These processes produce a Smaller Data Size than the Original File was uploaded to the Walrus Information Store, but also offer a Durable and Restorable Storage Method.
(Decentralization & Storage Nodes) Each storage node is an independent server or validator that is tasked with accepting and storing portions of data. The responsibilities of all storage nodes include: (1) Receiving encoded slivers of data (2) Responding to storage availability challenges (3) Providing information related to stored data when requested by users (4) Providing users with incentives in the form of fWAL WAL honest activity as a storage validator.
Storage nodes operate via delegated proof-of-stake (dPoS) which provides users the ability to stake WAL in order to promote good behaviour from the storage nodes. Through this process, users have a vested interest in ensuring that their selected storage providers are maintaining consistent levels of reliability.
(What the Blockchain Stores) It is important to note that Walrus does not store the actual pieces of data on the Sui BlockChain. Rather, the Sui BlockChain stores:
(1) Metadata associated with the data blob
(2) Proofs that indicate that the data blob is still available to users
(3) Details regarding who owns the data blob, and when it expires
In other words, the Sui BlockChain is a decentralized record-keeping system that tracks contractual relationships between users and their storage providers. The Sui BlockChain operates as a verifiable ledger for tracking how much data is stored across all storage nodes in the Walrus Network.
. Retrieval of Data
When an individual wishes to access a file:
Data from the Storage Nodes will be collected to create an aggregate node’s assembly of slivers representing the needed data.
These slivers would then be assembled back together to recreate the original file.
You can access the data quickly through the CDN (Content Delivery Network) or cache.
This model for storing data "Off-Chain"/"Storage Nodes" and containing cryptographic proof on the blockchain with "On-chain Proofs" allows for quick and simple access to your file without needing to interact with the blockchain.
📍 Where Does Walrus Store Data Geographically & Architecturally?
There are three primary layers where Walrus stores its data:
🔹 (1) On Decentralized Storage Nodes
Walrus’ Slivers are stored on multiple decentralized nodes around the world. The Decentralized Storage Layer consists of:
📍 Public Node Operators
📍 Staked Validators
📍 Independent Hosting Companies
Geographic decentralization of Walrus’ Slivers ensures a reduced likelihood of censorship and/or downtime.
🔹 (2) On the Sui Blockchain (On-chain only Metadata)
What is stored on-chain?
✔ Blob Identifiers
✔ Proofs of Availability
✔ Ownership and Scheduling Information
All of the actual Blobs are not stored on the blockchain; however, the metadata and Proofs of Availability are available on-chain.
3. CDN/Cache Page
Blurbs accessed often can be stored in a temporary cache on CDN Nodes to increase our user’s performance. This Cache Layer is optional and will not impact Security or Decentralisation in any way.
📈Walrus Data Storage Visualisation
Below are two visualisations to help visualize the way in which Walrus stores and retrieves blob data.
📊Graph — Blob Storage Procedure

Tally bar - Storage distribution

Token Mechanics & Storage Payments
The Walrus Token is the WAL token and it will do the following:
💰Pay for storage.
💰Support node operators and stakers.
💰Allow governance.
Stakeholders will be able to pay for certain amounts of epochs in WAL and over time, payments will be paid out to storage nodes.
walrus.xyz
Example Use Cases: Walrus has a few strong use cases that can be described below.
1) Storing large files - For developers, they can store 1GB or larger files (e.g. movies) or datasets without paying traditional cloud costs.
2) AI and ML - Users can store their training data, model weights, and checkpoints with proof of verifiability of availability.
3) Media and NFTs - All NFT media should not be hosted on any one location, rather they should all be hosted in a distributed way with proof of verifiability of the asset.
4) Decentralized Front Ends - The entire website and all components will be hosted in a decentralized manner through the Walrus Ecosystem.

📌 Importance of Walrus
Data storage and availability are major bottleneck issues within decentralized environments. Walrus solves these problems by utilizing a combination of:
💡 Efficient encoding methods
⚙ Proofs that are anchored to the Blockchain
📦 Distributed scalable storage
🏦 A WAL token based economic incentive model.
Due to this combination, Walrus becomes not only a way to store data but also provides a foundation for the development of DApps, Data Marketplaces and Programmatic Data Assets.
As decentralized ecosystems grow, services like Walrus will be essential infrastructure — just as cloud storage is today.
Conclusion
Walrus is a game changing development in the industry of storing files via a distributed approach to storing files. Walrus intelligently separates both content from verification. Files are kept off-chain, while ownership and proof of ownership are recorded on-chain. This ‘hybrid’ approach will ensure reliability, efficiency, and decentralization.
Бinance Academy
Summarize:
✔ The data is fragmented and saved to numerous different nodes,
✔ All metadata and proof exist on the Sui blockchain,
✔ The files can be recreated by gathering all the pieces back together,
✔ The $WAL Token is used for payment and reward purposes.
@Walrus 🦭/acc #walrus

One reason why Walrus is so significant is that cryptocurrency continues to create more records of information than current Blockchains are capable of handling, so we need a repository or storage solution that allows us to safely store all this information for the traders. Traders look at the storage portion of the technology as an "Infrastructure Play; however, Walrus is more than just a generic Storage Network. It is a platform developed to handle large sets of information, including files, multimedia, AI datasets, and even Application State. Walrus is also designed to allow the aforementioned types of data to become verified, priced, and consumed/utilized within OnChain Applications instead of simply storing the data "somewhere" or losing it forever. The choices we made when designing Walrus may seem to be more technical in focus; however, they have a major impact on how far our overall Ecosystem will be able to grow without forcing every single Blockchain to become a very costly hard drive.
Initially, Walrus was a project developed by Mysten Labs, with the team being closely associated with the early development of Sui. Now, however, they are operating independently from Mysten Labs and have developed a new token, WAL, and have established their own foundation. This independence is important because traders often do not realize how much of a difference it makes for these incentives. When a network becomes an entity in and of itself, it has the ability to attract those who want to be involved as the network expands. In addition to node operators and builders, companies who want to have a storage layer that is not controlled by one company's development plan will have the opportunity and incentive to develop relationships with the Walrus Network. The Walrus Network has clearly defined plans within their white paper regarding the delegation of proof-of-stake and that WAL will be the main currency used for governance and operating the Walrus Network through its storage nodes.
To better understand what Walrus does differently than the original version, we can look at Walrus as an alternative approach to the frequent practice of replication (i.e. creating multiple identical copies of a file on separate servers). Traditional decentralized storage (including BlockChain) relies heavily on the practice of creating duplicate copies to provide fault tolerance, however, replication has both the benefit of being simple to do but it can also incur a lot of costs to store.
Erasure coding is a method that Walrus is using that was designed specifically for larger blobs (or larger files) so that Walrus can take one large file and "erase" middle sections of it and store it on 9 or 10 nodes, and still be able to recover it even if 1 or 2 of those nodes go down.
In the whitepaper, one of the proposals is aimed at maximizing storage resiliency while minimizing storage overhead (as compared to replication). The cost of storage economics is the main reason why the most successful storage networks are able to achieve widespread use versus just remaining niche.
For investors and traders alike, the main emphasis will be on the fact that “Will the demand for the Token be driven by the technology?” and “Does the Technology create a utility for the Token in actual use?” The WAL Token (WAL) has been created to provide an economic benefit to the Protocol in that WAL is being used not only by users for storage payments but also by stakeholders for incentivization to operate the storage nodes, stake and contribute to governance of the Network. According to the Protocol's own Token Utility page, WAL will also serve as an incentive to users and represent the true economic value of the Protocol. In layman's terms, if Walrus gains traction as a true-use Platform, WAL will be incorporated into the Flow of Value: Users will PAY to store and retrieve, Node Operators will receive REWARDS, and STAKERS will help to maintain and secure the Network from attack. The key "long-term" opportunity is that Walrus will not merely be about launching a Token but rather building a sustainable Market for Data Storage and Data Retrieval.
The Walrus has become very popular with traders, in part because of the timing and because Walrus provides a necessary real-world solution to the exploding amounts of data from AI, media, applications, and on-chain that want to create better user experiences. Walrus defines itself as a place where "data can be both trustworthy, useful, and able to be governed," thereby projecting confidence in the way future apps will treat the ownership and ability to verify data, not just as an afterthought. Additionally, the Sui platform is also suitable for the Walrus because it offers high-throughput and low-latency solutions for storing data that can create a more cost-effective and efficient network of data storage at a larger scale.
The Mainnet is important for other reasons too; it represents when the theoretical market ends. Walrus' Public Mainnet Launch was announced March 27, 2025, with an emphasis on "Programmable Storage" (the ability to interact directly with Stored Data rather than just a dump location). The next test for an Infrastructure Token is on its treatment after mainnet launch; how did it continue growing with activity? Did your developers actually build out long-term utility/use for the Token? Walrus also cited testnet wipes occurring at various intervals as normal and healthy events for rapidly evolving infrastructure projects. The conclusion for developers who want to find stability should build off of mainnet.
Walrus has caught the eye of investors by completing a major token sale for over $140 million, sending out press releases for its token sale ahead of mainnet launch indicating it has attracted a number of serious investors, which would signal ample resources to develop its ecosystem. At the same time, there have also been a number of questions surrounding the typical pressures placed on investors regarding when their tokens will be available and the manner in which those tokens will be distributed. Serious traders should not concentrate solely on the "headline" of this token sale but also keep close tabs on the "schedules" associated with token unlocks, as well as the potential emissions associated with each time period. As with many storage networks, rewards attract more nodes than actually needed - as such, there is a continual influx of new nodes, but unlock-related selling will cap the token's performance until organic demand is sufficient to support further growth.
Walrus is important because the crypto industry generates more data than all blockchains combined can handle, so data must therefore exist in a "trustworthy" place. Many traders view storage solutions as infrastructure play, but Walrus is not just another generic storage solution. It was built as a solution to manage large amounts of data - not just a "generic" data storage network; it specifically addresses file/media/AI-related data and application states and allows developers to utilize that data as verified and priced components of their on-chain applications rather than as something that is simply parked in one place and then ignored. Although it may sound like a technical design decision, this unique development will create many opportunities for larger ecosystems to grow and develop without all blockchains having to become incredibly expensive hard drives.
Mysten Labs introduced Walrus as a project, which was then developed further into a geographically-diverse, independent decentralized network with its own token, WAL, and its foundation. This matters, as traders tend to assume that the level of "independence" of the network will play a major role in influencing incentives. Once a network has become independent and has its own identity, it will be much more attractive to Node Operators, Builders, and Partners who desire a storage layer that is not reliant on one company's designated strategy. The official white paper has announced.
I have a pretty simple view on the Walrus project: it will seem pretty boring day to day until someone understands how to leverage Walrus' win in building infrastructure for the future.
Think of the Storage primitive, while it is not particularly glamorous today, if eventually accepted as the standard default primitive by all building projects, Walrus will very quickly become a fundamental requirement of its ecosystem.
In my opinion future success of Walrus will depend on whether Walrus achieves "normal" status for the future and building infrastructures of its ecosystem and whether the design of Walrus' tokens stays harmonized with and reflects the value of actual usage, versus creating short-term economic incentives for large amounts of buyers.
If Walrus achieves its goals, it will not happen because of storytelling, it will happen when Walrus serves as the anchor for future developments of infrastructure and when the users on Walrus continue paying for storage without a second thought.
When these conditions are met, that is when the gravity of infrastructure will truly deliver what developers and users desire.
@Walrus 🦭/acc #walrus $WAL

Moje pochopení důležitosti „Walrus“ nevzešlo z diskuse o jeho ceně nebo dlouhého vlákna o trhu, ale spíše z jednoduché, avšak univerzální výzvy, kterou každý, kdo vytváří aplikaci AI nebo vlastní NFT, čelí: Jak chránit data (soubory obrázků, atd.), pokud někdo najednou získá kontrolu nad nimi a už neumožní přístup k vašim souborům? I když to zní jednoduše, Walrus se již současně používá v oblasti AI a NFT právě z tohoto důvodu. Walrus je decenteralizovaný způsob ukládání velkých neštrukturovaných souborů (např. zvuk, video, PDF, obrázky atd.) a využívá Sui jako vrstvy koordinace, která řídí fungování jejího úložného sítě. Cílem projektu je poskytnout spolehlivý, hodnotný a říditelný přístup k datům ve světě, nikoli jen k tomu, aby data byla někde hostována.