2026-01-11 04:27:23

While AWS and Google Cloud are still dividing the global storage market, a Sui-based project is changing the game. Walrus tells a convincing story with three sets of data: storage costs are only one-fifth of traditional cloud services, it has already hosted 37% of Sui ecosystem NFT metadata, and after the mainnet launch, its market value once surged to $1.2 billion.

The underlying principles are actually not complicated. Walrus adopts a two-layer architecture—Sui chain handles metadata management, while off-chain distributed nodes perform actual file storage. The key is the RedStuff encoding algorithm, which divides files into multiple fragments. Even if some nodes fail, the system can still fully recover the data. It sounds like traditional distributed backup storage, but with blockchain incentive mechanisms, competition automatically drives prices down.

Why are AI companies eager to use it? Simply put, AI training processes massive amounts of unstructured data—images, speech, text. Storing this on traditional clouds is expensive and poses privacy risks. Walrus’s approach is different: enterprises can upload and call data anonymously, and through programmable storage smart contracts, they can establish ownership, turning data from a pure cost into an income-generating asset. For example, when collaborating with Bittensor to store distributed training datasets, storage fees are determined by node bidding, and the system automatically takes the 66.67th percentile bid to prevent monopolistic pricing.

But things aren’t perfect. Early investors control 30% of the initial nodes, raising concerns about centralization. Additionally, allowing data deletion conflicts somewhat with the blockchain’s "permanent storage" principle, and regulatory considerations still need to be addressed. However, the trend is shifting—Morgan Stanley is already testing storing KYC compliance documents on the chain, and the entry of traditional financial institutions could be the next trigger.

Walrus’s ambition isn’t to replace all cloud storage but to carve out dedicated channels in high-frequency, high-privacy-demand fields. The demand for low latency and data sovereignty in AI, finance, and decentralized applications gives it opportunities. Plus, with cross-chain compatibility expanding (already supporting Ethereum, Solana), if it continues to develop, it could truly become the "water and electricity grid" of the Web3 data layer.

SUI0.25%

WAL3.31%

TAO2.04%

ETH0.84%

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

9 Likes

Reward
9
4
Repost
Share

Comment

0/400

PanicSeller

· 10h ago

Early nodes concentrating 30% really can't hold up, this is a common problem in Web3 projects... But the cost indeed can't match one-fifth of that number, there's something there. By the way, Walrus and Bittensor's combined efforts, if they can really cut down AI data costs, competing with AWS and those guys would be awesome. I'm just confused about the contradiction of "permanent storage" deleting data; solving this regulatory challenge is the real test. Morgan Stanley entering the KYC space... will it just be traditional finance testing the waters and then pulling out? Let's wait and see.

View OriginalReply0

NFTBlackHole

· 10h ago

Whoa, Walrus's node centralization this time really is a hidden risk... Early large holders controlling 30% of the initial nodes, isn't that just a different name for centralization? That said, the cost being one-fifth is indeed tempting. Automatically fetching the 66.67 percentile price for storage fees—this anti-monopoly move is interesting, but the problem is the conflict between data permanence and deletion functionality. Regulation definitely needs to be watched. Is anyone actually using it, or is it another storytelling project for fundraising? --- Walrus is hot, but I still feel that the 30% of nodes held by early investors might cut retail investors... It really depends on how the subsequent node distribution evolves. --- Chasing a 1.2 billion market cap? That number sounds impressive, but who knows about actual usage and token liquidity... Whether AI companies use it is the key. --- Morgan Stanley has started testing on-chain storage of KYC files. If this truly breaks through traditional finance, then Walrus might really have a shot.

View OriginalReply0

GasFeeSurvivor

· 10h ago

To be honest, Walrus's gameplay really hits my spot. The cost of one-fifth of this number alone is worth paying attention to. However, the centralization aspect is still quite concerning. Early investors holding 30% of the nodes... isn't this just the old trick? Morgan Stanley testing the waters is reliable; the true entry of financial institutions is a sign of breaking the deadlock. Permanent data storage vs. allowing deletion—this is indeed contradictory. It depends on how it will be reconciled later. Cross-chain support for Solana and Ethereum is also good; it all depends on whether it can truly support large-scale applications.

View OriginalReply0

BearMarketHustler

· 10h ago

Damn, Walrus's move is really impressive. One-fifth of the cost? It was long overdue for someone to smash the traditional cloud's reputation. However, 30% of the nodes are in early investors' hands, so this risk definitely needs to be considered. Under the guise of decentralization, there's still a bit of centralization feeling. Wait, Morgan Stanley is also getting involved? This could really turn into a breakout moment for traditional finance. The Sui ecosystem is now so hot, we need to keep a close eye on whether node expansion will tell another story.

View OriginalReply0