Most confidential systems release secrets as soon as they get asked.
Nox flips it. A Virtual Machine must first prove it booted the expected operating system and application stack.
Only then does it get access to keys.
Boot time verification isn't paperwork. It's the gate.
10·CNeutral
H
Hanami3h agonews
Most confidential systems release secrets as soon as they get asked.
Nox flips it. A Virtual Machine must first prove it booted the expected operating system and application stack.
Only then does it get access to keys.
Boot time verification isn't paperwork. It's the gate.
55·BLong
o
oi_change4h agomarket
TAC/USDT OI 5min Up 5.50% $6.11M rose to $6.44M, Price 5.16%, New Longs Entering
0·-Neutral
m
meme22h agomeme
Goldman Sachs expects hyperscalers to spend $1.14 trillion on AI data centers in 2027, and roughly a third of that gets borrowed rather than earned.
Some of that borrowing has already moved onto crypto rails (save this).
Those same firms spent $405 billion in 2025, and they're tracking about $750 billion this year, so the bill nearly triples in two years. Growth is expected to cool off, from roughly 85% this year to about 52% next year, but that's still the fastest spending ramp on record.
And the count only covers the hyperscalers. Goldman tallied another $412 billion of AI-linked borrowing this year from chipmakers and data center builders. The market is laser focused on the assets/sectors where that $1T+ is headed, which is Nvidia, memory and power. Very few people, by comparison, are watching the stack of lenders who front that cash.
The first in the stack is the investment-grade bond market.
Hyperscalers sold $108 billion of those bonds in 2025, about 26% of their capex. They sold $194 billion in the first half of this year alone. Goldman is expecting $400 billion against 2027 capex.
Alphabet posted its first negative free cash flow quarter since 2004 in July. Meta handed BlackRock an 80% stake in its El Paso site, with $12.5 billion of debt behind it.
Second in the stack is private credit, and it exists because those bonds only serve borrowers rated AAA or AA. Goldman says project finance and asset-backed lending are the channels soaking up the rest. The same five firms have already disclosed about $1.2 trillion of lease commitments.
Third in the stack is where you come in, because smaller GPU operators can't reach either of the first two.
USD(dot)AI lends against Nvidia hardware tokenized as collateral. By its own count it originated its first $100 million of GPU-backed loans in Q1, and the yield comes from borrower repayments rather than token emissions.
Figure runs that same plumbing at real size. It reported $4.3 billion of consumer loan volume in Q2, up 132% on the year, with 65% of it now crossing its tokenized marketplace.
$1.14 trillion has to get borrowed from somewhere, and it splits across bonds, private credit and onchain rails.
We're tracking all three, and the onchain one is barely off the ground.
Don't miss our next entry.
Follow us for more @MilkRoadDaily, and track each one of our investments in real time with Milk Road PRO (prices go from $25/m to $39/m next Weds): link.milkroad.com/zwu5a9
75·ALong
m
meme22h agomeme
HBM MASSIVE CONTENT DOWNGRADE ALERT🚨🚨: Nvidia Rubin Ultra will ship with 192GB of HBM4 8-hi. Not only is this a huge downgrade from the original Rubin Ultra which was previewed with 1TB of HBM, but it is even lower than regular Rubin which has 288GB. How did HBM content for Rubin Ultra get cut to 1/5th of the original? The number of cubes got cut to 8 when 4 die Rubin Ultra was scrapped. Then the stack height got reduce to 8-hi instead of 16-hi. Last it is cut to HBM4 which uses 24Gb dies instead of 32Gb dies with HBM4E, although an 8-hi HBM4E upgrade could com later. What was Nvidia's reasoning behind this? (1/2)🧵
75·AShort
m
meme22h agomeme
HBM MASSIVE CONTENT DOWNGRADE ALERT🚨🚨: Nvidia Rubin Ultra will ship with 192GB of HBM4 8-hi. Not only is this a huge downgrade from the original Rubin Ultra which was previewed with 1TB of HBM, but it is even lower than regular Rubin which has 288GB. How did HBM content for Rubin Ultra get cut to 1/5th of the original? The number of cubes got cut to 8 when 4 die Rubin Ultra was scrapped. Then the stack height got reduce to 8-hi instead of 16-hi. Last it is cut to HBM4 which uses 24Gb dies instead of 32Gb dies with HBM4E, although an 8-hi HBM4E upgrade could com later. What was Nvidia's reasoning behind this? (1/2)🧵
75·AShort
m
meme8/24meme
quote: Pat sums up the agentic finance thesis well.
MANTRA, NVNM and DGML are combining forces, all powered by Inveniam, to bring RWAs and private markets onchain with verifiable, AI ready infrastructure. https://x.com/1438480878687166473/status/2091483453740957774 | Delivering systematic trading of private markets is what we have been working on for years, and it is coming and coming fast, (see WSJ article, https://t.co/s605VbOc3K).
Fundamentally these are document based markets, this is why we have focused so much on unstructured document based data. A public equity algorithm can trade against standardized prices, financial statements, corporate actions, and market data. A private-market strategy has to interpret PDFs, operating agreements, loan documents, appraisals, rent rolls, financial statements, cap tables, covenants, and servicing reports. That makes systematic trading extraordinarily difficult. This is now going to change.
A. Tokenization makes private assets tradable.
B. Inveniam makes their data usable.
C. https://t.co/3jK7WUPHhm makes their documents machine-readable.
D. MANTRA makes the resulting transactions and proofs executable on-chain.
Together, they create the infrastructure required for systematic trading in private markets.
The architecture is what changes this:
1. Tokenization creates the instrument.
Private Credit interval funds, CRE, Infrasturcture, or Private Equity funds, all the way down to a unique large asset such as a building, private credit position, individual fund interest, infrastructure asset, or other private-market exposure can be represented digitally. Ownership, transfer restrictions, distributions, and other economic rights become programmable. But tokenization alone doesn't solve the problem. A token without trustworthy underlying data is simply a more efficient wrapper around an opaque asset.
2. @Inveniam creates the private-market data layer.
Inveniam connects the token to the actual operating and financial information underlying the asset. Instead of sending copies of sensitive information everywhere, the asset owner maintains control while data can be indexed, permissioned, queried, and used by machines. Cryptographic proofs can establish origin, state, and process—where information came from, what its state was at a particular point, and what happened to it. Inveniam is the control plane for private market assets data.
3. https://t.co/3jK7WUPHhm turns documents into computational objects.
This is particularly important because much of the information that determines private-asset value lives inside documents. DGML can transform those documents from blobs of text into consistently tagged, typed, attributable information. An agent doesn't merely read "this loan agreement." It can identify the , , , , or other domain-specific elements and trace them back to their source. That starts making unstructured private-market information queryable like structured market data.
4. @MANTRA provides the transaction and settlement layer. Once the asset is tokenized and the information supporting it is machine-readable and verifiable, MANTRA can provide the blockchain environment in which ownership, permissions, transfers, proofs, and settlement can interact. Importantly, the sensitive underlying documents don't have to live publicly on-chain. The chain can carry the asset, transaction state, and cryptographic evidence necessary to establish that off-chain information satisfies defined conditions.
That gets us to systematic private-market trading.
A machine can now move through something analogous to:
Asset → Documents → DGML → Inveniam Data Layer → Verified State → Models/Agents → Trading Signal → Compliance Rules → MANTRA Transaction → Settlement → New Verified State
Consider a commercial real-estate credit strategy. Instead of an analyst manually reviewing hundreds of loan files, an agent could ask:
Find every eligible tokenized CRE loan where LTV < 60%, DSCR > 1.5x, occupancy > 90%, no covenant breach exists, the underlying data was refreshed within 30 days, and the expected yield exceeds 9%.
DGML gives the machine semantic understanding of the documents. Inveniam supplies the permissioned data and evidence supporting those values. The model calculates the opportunity. MANTRA determines whether the instrument can legally and operationally move between the relevant parties and provides execution and settlement.
And then something much more powerful becomes possible. The system doesn't merely trade assets; it can continuously monitor them.
If occupancy falls, a covenant is breached, an appraisal changes, a borrower refinances, a property produces new operating data, or a document changes, the verified state changes. That can automatically change valuation, risk scoring, collateral eligibility, portfolio weighting, or ultimately generate a buy/sell signal.
The four layers very simply:
1) Tokenization - Makes the asset programmable and transferable
2) DGML - Makes the documents machine-readable
3) Inveniam - Makes the underlying data permissioned, attributable and verifiable
4) MANTRA - Makes ownership, proofs, execution and settlement programmable
The end product isn't really "tokenization."
It is machine-readable private markets.
Public markets became systematic because securities, market data, financial information, execution, and settlement became standardized enough for machines to operate against them.
Private markets haven't had that infrastructure.
Tokenization + DGML + Inveniam + MANTRA is an attempt to build that missing stack.
Inveniam + DGML + MANTRA can do what both Bloomberg and Exchanges do for public markets for private markets—turning documents into verified data, data into signals, signals into transactions, and transactions into continuously verifiable asset states.
That is the path from private-market digitization → tokenization → machine-readable assets → systematic trading → agentic private markets.
Now I will add NVNM as a fifth layer—the layer that turns agent activity from a black box into provable economic activity.
5. @NVNM — The Proof and Accountability Layer for AI Agents
Once AI agents begin analyzing, valuing, underwriting, monitoring, and ultimately trading private-market assets, the critical question becomes:
How do we know why the agent made the decision? We need cryptographically secure evidence.
In public markets, systematic strategies rely on standardized data feeds and established market infrastructure. In private markets, an agent may make a decision using dozens of documents, proprietary datasets, calculations, models, and intermediate transformations. For institutions to allow agents to deploy capital, the agent needs to be able to show its work.
That is where NVNM fits.
NVNM provides the mechanism for an agent to cryptographically prove three things about the information behind a decision:
Proof of Origin — What did the agent know? The agent can demonstrate which source documents, datasets, counterparties, systems, or other inputs were used without necessarily exposing the confidential underlying data.
Proof of Process — What did the agent do? The system can attest to the workflow, models, calculations, transformations, policies, and approved processes applied to those inputs.
Proof of State — What was true when the decision was made? The agent can prove the state and version of the underlying information at the exact point in time when it generated a valuation, recommendation, risk determination, or trade.
The proofs—not the confidential data itself—can then be anchored on-chain.
The result is an immutable evidence trail connecting data → analysis → decision → transaction.
That changes the nature of an AI-generated trade. Instead of:
Agent says: BUY
the market can receive:
BUY + verified inputs + verified process + verified state + provenance + timestamp + authorization.
This becomes particularly powerful when multiple agents participate in the same market. An underwriting agent could establish asset eligibility. A valuation agent could calculate fair value. A risk agent could determine portfolio exposure. A compliance agent could approve the counterparty and transaction. An execution agent could then submit the transaction to MANTRA.
Each agent can produce NVNM proofs of the work it performed.
So the complete architecture becomes:
PRIVATE ASSET
↓
DGML — What does the asset's documentation mean?
↓
INVENIAM — What data can the agent securely use?
↓
AI AGENTS — What should we do?
↓
NVNM — Prove what the agents knew, what they did, and what was true.
↓
MANTRA — Execute and settle the resulting transaction.
↓
NEW VERIFIED STATE
↓
Repeat
This creates a highly virtuous feedback loop. Every transaction produces a new state. New operating information produces another state. Agents continuously evaluate those changes, create new proofs, and potentially generate new transactions.
The Five-Layer Systematic Private Markets Stack
1) Tokenization - Digitizes ownership and economic rights - Tradable private assets
2) DGML - Makes documents semantically machine-readable - Computable private-market information
3) Inveniam - Permissioned access, indexing, attribution and verification - Institutional-quality data infrastructure
4) NVNM - Proves agent origin, process and state - Auditable AI decision-making
5) MANTRA - On-chain asset, transaction and settlement infrastructure - Programmable markets
This sharpens the overall thesis considerably:
We aren't simply tokenizing private assets. We are building the infrastructure that allows machines to understand them, value them, prove why they acted, and transact them.
The critical difference is proof and data sovereignity.
AI can already read a document, while leaking that data. AI can already build a model, while stealing your data. AI can already generate a trading recommendation.
The institutional problem is whether a fiduciary, regulator, auditor, investment committee, lender, or counterparty can determine what information the agent relied upon, where that information originated, what process was applied, what the information's state was when the decision occurred, and whether the agent was authorized to act.
NVNM turns the agent's answer into a provable answer.
That is ultimately what could make systematic trading of private assets institutionally viable:
Machine-readable assets.
Machine-readable data.
Machine-executable markets.
Machine-generated decisions.
Machine-verifiable proof.
Or in summary:
DGML lets the agent understand. Inveniam lets the agent access and organize. NVNM lets the agent prove. MANTRA lets the agent transact. Tokenization gives it something to trade.
That is the foundation for systematic—and eventually agentic (autonomous)—private markets.
The moment everyone has been waiting for is finally here. The token allocation details have been announced, so make sure to head over to the tweet below and check your allocation. If you're eligible, don't miss the opportunity to claim your tokens and complete the process.
This is a major milestone for the community, and it's exciting to see everyone finally reaching this stage after all the support, patience, and dedication along the way. Whether you've been following from the beginning or joined during the journey, this is an important moment to be part of.
Make sure to carefully read the details in the allocation tweet, follow the official steps, and stay updated on any further announcements. Avoid missing important information and double-check everything before completing your claim.
Congratulations to everyone who is receiving an allocation. This is another big step forward, and the future looks exciting as the community continues to grow. Thank you to everyone who has supported the project and helped build the momentum that brought us here.
Go check the allocation tweet down below and see your token allocation. 👇 @Mitchel90886406 @stack__sats @OmRawat_ @dB00085 @hbeckeri @Sandblaster80 @metagammar @MorenaSatoshi @emilioascencioo @alfredwingman @chipdapotato @presence908 @Zoneveldd @ongJin1988 @0xPowery @Mattymorgan1989 @Olarewajhu @GTK_xt @QuickQuackSol @trapicidio @aspirevs @Brian15725920 @GlaucoMaiaGFM @wake06_ @NishantaD @liwei4021 @CL_CLACL @Donclairo @fiatpeasant @VelisNFT @CallMeThat8 @rexamphetamine @cointur2 @kelz_2682 @mostlybedlam @DGCM260076 @Nill__kiggers @B33_lay @0xCarterOnChain @Ry_Guy_NFT
65·B+Long
m
meme8/23meme
quote: The AI race is moving down the stack.
Models → Compute → Energy.
The next layer to be rebuilt?
Capital.
When AI can reason and act autonomously, capital infrastructure can no longer be designed only for humans.
Machine-native capital markets are coming. | BREAKING: Nvidia, $NVDA, has agreed to provide a more than $100 billion backstop for a massive new OpenAI data center in Ohio, per FT.
Details include:
1. Nvidia will provide credit support for the "land, power and shell" for the facility capped at $105 billion
2. The data center is being developed alongside a SoftBank-led energy companies
3. Nvidia will also invest $1.5 billion into SB Energy, an energy company founded as part of the SoftBank Group in 2019
4. OpenAI plans to lease as much as 8 gigawatts of AI computing power at the data center in Pike County, Ohio, which will take until 2032 to complete
This will mark one of the largest AI infrastructure deals yet.
80·ALong
m
meme8/22meme
quote: The objective of @USDai_Official is to create a financial platform for the next 10,000 neoclouds
The GPU mortgage is just the start, afterwards it’s settlement, and then it’s the entire financial stack as all dollars in AI does one thing only: pays down the mortgage https://x.com/dylan522p/status/2090957403424776270 | God damnit, every one of my AI founder friends who actually have revenue are now just transforming into neoclouds with value add on top
70·B+Long
m
meme8/21meme
Tankers stack up as #Venezuela sells oil faster than its ports can handle #oott https://www.reuters.com/business/energy/tankers-stack-up-venezuela-sells-oil-faster-than-its-ports-can-handle-2026-08-21/
90·A+Long
n
news8/20news
A milestone for our infrastructure: our first NVIDIA Vera Rubin racks are here and now running our training stack. This is an important step as we expand compute that powers OpenAI's next generation of frontier AI pre-training. https://t.co/HXk6DrfLav
75·ALong
m
meme8/20meme
A milestone for our infrastructure: our first NVIDIA Vera Rubin racks are here and now running our training stack. This is an important step as we expand compute that powers OpenAI's next generation of frontier AI pre-training.
80·ALong
T
Twitter8/20news
Zcash: Financial Privacy in the Age of AI
Key Takeaways New technologies can upgrade the financial system but may also reduce financial privacy. This dynamic emerged with the digitization of banking and again with the rise of the internet. Today, the expansion of AI may create similar challenges. Zcash @Zcash is a decentralized digital currency much like Bitcoin, but with built-in privacy features. In Zcash private transactions, the sender, receiver, and amount are all obscured, allowing Zcash to function like physical cash (which was Bitcoin’s original vision). In a world of AI-powered financial surveillance, these features could become essential. Zcash is nearly 10 years old but may be entering a new chapter. Use of its shielding technology is increasing, and new capital is entering the ecosystem to support wallet development and Zcash mining. Zcash’s ZEC token is valued at just ~$8 billion, or 0.6% of the Currencies Crypto Sector. If it were to capture just 5% of this market segment, its value would be ~9x greater. Zcash is a lower market cap crypto asset and should be considered higher risk. Privacy is not a niche feature of money; it’s part of what makes money work. People don’t want every transaction, balance, and financial relationship exposed. Cash has this property. And historically, intermediated (bank-based) financial systems have preserved a level of practical confidentiality. They limit third-party access to financial records, though information can still be exposed. Even law enforcement generally requires a court order to access personal financial data. Technological and social change can put pressure on these systems, often triggering public debate about financial privacy and new methods to protect it. In the 1970s, the digitization of financial records and the Bank Secrecy Act brought more attention to financial privacy (Exhibit 1). Similarly, in the late 1990s and early 2000s, the expansion of the internet and online banking, along with the Patriot Act, led to another wave of focus on financial privacy—and eventually broader use of encryption, 2FA, and related tools. Exhibit 1: Technological change can bring focus to financial privacy
Today, Grayscale Research believes we are on the cusp of a third wave of widespread public attention on financial privacy, driven by the growth in stablecoins and other blockchain-based applications, as well as artificial intelligence (AI), which may offer new methods of financial surveillance. Based on higher-frequency indicators, public attention on these issues already seems to be increasing (Exhibit 2). Exhibit 2: Recent increase in focus on financial privacy
This is where Zcash comes in. Zcash is a decentralized digital currency much like Bitcoin, but with sophisticated privacy features. Bitcoin made digital scarcity credible, but not digital privacy. Its transparency supports auditability, but limits money-like use cases where confidentiality matters. If cryptocurrency users eventually come to value private digital money as distinct from transparent digital assets, Zcash could capture part of that demand. At roughly 0.4% of total cryptocurrency market cap, ZEC appears small, in our view, relative to the quality of its technology and the potential addressable market of privacy-preserving digital money. Privacy as a Product Category On public blockchains, balances, counterparties, and transaction histories can remain visible indefinitely. Even when users transact through pseudonymous addresses, activity often becomes linkable over time—through exchanges, counterparties, wallet behavior, and blockchain analytics. In practice, these systems often reveal more information than users expect. However, most users expect some level of financial privacy. Individuals may not want balances or spending history exposed by default. Businesses may want to keep suppliers, payroll, treasury flows, and customer relationships confidential. Institutions may not want wallet structures or transaction patterns mapped in real time. Privacy demand is therefore not limited to users seeking full anonymity. In many cases, it reflects ordinary preferences around confidentiality in economic life. Yet, within crypto, the privacy category’s main issue is that it has often resulted in weaker distribution. Stronger privacy protections can create more friction in exchange support, wallet integration, and market access. As a result, privacy in crypto has not just been a technical feature; it has also involved a commercial tradeoff. Approaches to Onchain Privacy Not all crypto privacy systems hide the same information. Some mainly make transaction flows harder to trace on an otherwise transparent ledger. Others hide specific transaction fields more directly. Mixing-based privacy on transparent chains such as Bitcoin CoinJoin, Bitcoin Cash CashFusion, and Dash PrivateSend try to make transaction flows harder to trace, but the underlying ledger remains visible. Confidential transaction systems such as Litecoin MWEB can hide amounts and add structural privacy, while still differing from fully shielded systems. Default-private systems such as Monero use built-in privacy mechanisms to obscure key transaction details by default, rather than asking users to opt in. Shielded systems such as Zcash’s validate transfers without publicly revealing sender, recipient, or amount. Within this privacy space, Zcash stands out as an optionally private base-layer system whose shielded transfers can hide sender, receiver, and amount at the transaction layer. These features place Zcash in a different category from privacy overlays on transparent blockchains (Exhibit 3). Exhibit 3: Transaction privacy varies by cryptocurrency
What Zcash Is and How It Works Zcash is a public blockchain for peer-to-peer value transfer with a fixed 21 million coin supply and a Proof of Work security model. In many respects, it resembles Bitcoin. The main difference is that Zcash gives users the option to shield transaction data rather than publishing all transaction details by default. The network supports two transaction types. Transparent transactions are publicly visible onchain, similar to Bitcoin-style transfers. Shielded transactions verify that a transaction is valid without publicly revealing the sender, recipient, or amount. In effect, Zcash allows verification without full public disclosure. In Zcash, the privacy mechanism is a shielded pool validated with zero-knowledge proofs, which can reduce onchain disclosure substantially when users remain within shielded pools. Zero-knowledge proofs are a type of advanced cryptography that can be used for both privacy and scaling in a blockchain context. Zcash also supports selective disclosure through viewing keys, allowing users to share visibility into shielded activity with specific third parties when needed. This matters because shielded transactions are not just simply “opaque�; they are private from the public while still allowing user-controlled disclosure when needed. The Zcash Story: Old Protocol, New Chapter Zcash launched in 2016. This makes it old by crypto standards and easy to misread as a holdover from an earlier market cycle. But from the beginning, it was more ambitious than a simple focus on privacy suggests. The project grew out of a serious effort to bring zero-knowledge cryptography into a digital cash system. Its core idea was unique at the time: combine a Bitcoin-like monetary structure with the ability to transact privately on a public blockchain. That made Zcash one of the earliest attempts to build private digital cash as a base-layer asset, rather than as a wallet feature or privacy layer added after the fact. That ambition, however, ran ahead of the product infrastructure around it at the time. For much of its history, shielded functionality was harder to use than the transparent alternative because early shielded transactions were computationally demanding. Several major upgrades gradually reduced those constraints: Sapling (2018): made shielded usage materially more practical by sharply reducing proving time and memory requirements relative to Sprout4. Zcash’s own materials say Sapling payments could be constructed in as little as a few seconds and with only 40 megabytes of memory, a change explicitly tied to improving adoption. Orchard / NU5 (2022): modernized the cryptographic foundation by introducing the Orchard shielded protocol and Halo 2, removing the trusted-setup requirement for the new shielded pool. Ironwood (2026): restored supply integrity by introducing a new shielded pool built on the corrected Orchard circuit, preventing any hypothetical counterfeit ZEC from expanding the circulating supply. Unified Addresses / wallet improvements: helped reduce friction in wallet flows by making it easier for users to interact with both transparent and shielded receivers without relying on multiple address formats. The challenge, in other words, was not just building private transfers. It was making them easier to use at scale. That long effort to reduce technical friction may finally be paying off. One important piece of that story is the wallet layer. @zodl_co, formerly known as Zashi, is increasingly positioned not just as a place to hold ZEC, but as a shielded-first interface for using it. Current product materials highlight built-in swaps powered by Near Intents, CrossPay functionality that lets users send shielded ZEC while the recipient receives another asset, and support for cold storage of shielded balances. These features matter because they reduce the number of steps where users would otherwise leave the shielded context to fund a wallet, make a payment, or move into another asset. For Zcash, that’s important: the main bottleneck has historically not been whether private transfers were possible, but whether users could remain shielded through the full transaction flow. There are also signs of broader infrastructure support around the network itself. Foundry, one of the largest cryptocurrency mining pools, announced plans to launch a US-based institutional Zcash mining pool in April 2026, aimed at institutional and public-company miners. That does not solve shielded wallet adoption directly, but it does suggest a different kind of ecosystem maturation: more industrial-grade infrastructure, more operational support, and potentially more institutional legibility around the network. Finally, policy changes around fees and mempool behavior matter at the margin because they can make spam-style congestion less effective and improve network usability under load. That is not the core of the Zcash thesis, but it does fit the broader pattern: the network is being shaped not only for privacy in theory, but for more reliable use in practice. The next phase of the Zcash story is likely to depend on whether the protocol can move from “privacy works� to “privacy scales.� The key development trends to watch are: Tachyon, the clearest scaling proposal on the horizon. At a high level, Tachyon aims to reduce some of the cost and synchronization burdens that have historically made shielded usage harder to support at scale. Crosslink, an upgrade more about finality and network reliability than privacy itself. The idea is to strengthen settlement assurances without replacing Zcash’s existing Proof of Work structure altogether. For the investment case, the relevance is that Zcash’s future is not only about privacy features in isolation; it also hinges on whether the broader network becomes more robust and easier to trust operationally. Shielded assets, unlock beyond native ZEC. If revived, they could extend Zcash’s privacy model beyond ZEC itself and make shielded transfer functionality relevant to a wider set of assets or applications. What Is ZEC Worth? As of July 2026, the market is assigning very little value to privacy in crypto. Grayscale categorizes all the crypto assets with a “digital currency� use case in our proprietary Currencies Crypto Sector. Today, this crypto market segment includes 15 assets with a combined market capitalization of $1.4 trillion. Bitcoin is king of the category, with a market share of about 90%. Zcash’s ZEC token is valued at just ~$8 billion today, or 0.6% of the Currencies Crypto Sector (Exhibit 4). If it were to capture just 5% of this market segment, its value would be 9x greater. Exhibit 4: ZEC’s share of the Currencies market cap is small but growing
ZEC’s current valuation appears to reflect an assumption that privacy will remain peripheral: useful to some, but too narrow to command a meaningful premium. But if privacy becomes more valuable in a world of greater surveillance, tighter compliance rails, and more visible financial censorship, then a 0.6% market share may look less like a fair equilibrium and more like a market that has not fully priced the asset's optionality. The bullish version of the Zcash thesis is that the market is currently valuing ZEC as if privacy demand will stay contained, even though the chain already shows evidence that privacy usage is real. As of July 20, 2026, shielded transactions made up roughly 90% of Zcash transaction count, and shielded supply stood at about 4.2 million ZEC, or roughly 25% of current circulating supply (Exhibit 5). That suggests privacy is not merely a narrative attached to the asset, but an observable on-chain behavior. Exhibit 5: ZEC’s shielded supply share has grown to an all time high
In other words, Zcash does not need explosive growth today to justify potential upside; it needs the market to change its view about what privacy is worth tomorrow. If investors eventually decide that privacy should command even a modest premium inside crypto's monetary stack, ZEC's current valuation could prove conservative, because the market is still treating privacy as more an afterthought than a core property of money. Risks and Other Considerations Zcash’s investment case is not just a question of privacy demand. It also depends on how investors think about legacy cryptographic assumptions, future security risks, and the network’s ability to keep shipping major upgrades without breaking coordination. Regulatory risk Zcash’s regulatory risk is not simply that it offers privacy, but that regulators and service providers may disagree on whether its selective disclosure tools are sufficient for compliance. Zcash’s viewing-key model gives a clearer compliance story than systems with no disclosure mechanism at all: a Full Viewing Key/Unified Full Viewing Key can be shared with a chosen party to reveal incoming activity and, for standard outgoing shielded payments, recipient address, amount, and memo. But that disclosure is still permissioned rather than default public traceability, while global AML/CFT standards for Virtual Asset Service Providers continue to emphasize customer due diligence, recordkeeping, suspicious activity reporting, and the Travel Rule, and Markets in Crypto Assets (MiCA) adds authorization, disclosure, and supervisory requirements in the EU. For an investor, the right framing is that viewing keys likely reduce—but do not eliminate—the risk that exchanges, custodians, or regulators treat shielded activity as operationally or politically harder to support. Trusted-setup legacy pools One historical concern around Zcash has been the trusted setup used for the older Sprout and Sapling shielded protocols. NU5 introduced Orchard and the Halo proving system, which removed the trusted-setup requirement for the Orchard shielded pool. Separately, a soundness vulnerability was discovered in the original Orchard circuit in 2026 that could theoretically have enabled undetectable counterfeiting within the Orchard pool. A network upgrade (NU6.2) corrected the circuit, while Ironwood (NU6.3) introduces a separate shielded pool. By preventing new value from entering the legacy Orchard pool and requiring funds leaving it to pass through Zcash’s turnstile accounting mechanism, Ironwood is designed to restore users’ ability to independently verify the integrity of Zcash’s circulating supply. This materially improves the forward-looking cryptographic story. The residual issue is that older pools still exist, and legacy funds can remain in them. In practice, this is best understood as a diminishing legacy risk rather than the defining risk of the current shielded design, but it is still relevant when comparing older and newer parts of the protocol. Quantum risk Quantum computing is a long-horizon risk for most public blockchains, not a Zcash-specific problem. At a high level, transparent and shielded funds may face different exposure paths depending on which cryptographic components are visible onchain and when spend-authorizing keys are revealed. Electric Coin Company, a company historically associated with building and launching Zcash, had recent roadmap work explicitly including quantum resilience as an area of focus6, which suggests the issue is being treated as a real design consideration rather than a purely theoretical one. For an investment note, the most reasonable framing is that quantum risk matters, but likely on a longer time horizon than the near-term usability, adoption, and market-structure questions that dominate the Zcash thesis today. Execution risk Zcash has historically relied on major protocol upgrades to improve usability, security, and scalability. That creates execution risk in two forms. First, the roadmap itself is ambitious: proposals such as Tachyon, Crosslink, and related changes are meaningful technical efforts rather than simple parameter tweaks. Second, network upgrades require coordination across protocol engineers, wallet developers, infrastructure providers, and the broader ecosystem. Zcash’s upgrade process is structured and well documented through Zcash Improvement Proposals, which is a strength, but it also means the investment case depends in part on continued coordination and implementation quality. Conclusion Privacy has always been part of money’s usefulness, even if digital systems have made that easier to forget. Zcash was built around a limitation of transparent blockchains: they make transactions easy to verify, but difficult to keep confidential. After years of technical work, it now stands as one of the clearest attempts to close that gap. Whether private digital money becomes a major category or remains a niche one is still uncertain. But at current valuations, the market appears to be assigning little probability to privacy becoming materially more important from here. For investors, that may be the opportunity: not a bet that Zcash replaces Bitcoin, but that the value of private digital money has yet to be fully recognized.
Investments in digital assets are speculative investments that involve high degrees of risk, including a partial or total loss of invested funds. Investments in digital assets are not suitable for any investor that cannot afford loss of the entire investment. This information should not be relied upon as research, legal, tax or investment advice, or a recommendation regarding any products, strategies, or any investment in particular. This material is strictly for illustrative, educational, or informational purposes and is subject to change. This content does not constitute an offer to sell or the solicitation of an offer to sell or buy any security in any jurisdiction where such an offer or solicitation would be illegal. There is not enough information contained in this content to make an investment decision and any information contained herein should not be used as a basis for this purpose. This content does not constitute a recommendation or take into account the particular investment objectives, financial situations, or needs of investors. Grayscale Operating, LLC and its affiliates, directors, officers, and employees may have interests, financial or otherwise, in the digital assets and/or securities discussed. Certain of the statements contained herein may be statements of future expectations and other forward-looking statements that are based on Grayscale’s views and assumptions and involve known and unknown risks and uncertainties that could cause actual results, performance, or events to differ materially from those expressed or implied in such statements. In addition to statements that are forward-looking by reason of context, the words “may, will, should, could, can, expects, plans, intends, anticipates, believes, estimates, predicts, potential, projected, or continue� and similar expressions identify forward-looking statements. Grayscale assumes no obligation to update any forward-looking statements contained herein and you should not place undue reliance on such statements, which speak only as of the date hereof. Although Grayscale has taken reasonable care to ensure that the information contained herein is accurate, no representation or warranty (including liability towards third parties), expressed or implied, is made by Grayscale as to its accuracy, reliability, or completeness. You should not make any investment decisions based on these estimates and forward-looking statements. There is no guarantee that the market conditions during the past period will be present in the future. Rather, it is most likely that the future market conditions will differ significantly from those of this past period, which could have a materially adverse impact on future returns. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN. PAST PERFORMANCE IS NOT INDICATIVE OF FUTURE RESULTS. We selected the timeframe for our analysis because we believe it broadly constitutes the most complete historical dataset for the digital assets that we have chosen to analyze. All content is original and has been researched and produced by Grayscale Investments Sponsors, LLC (“Grayscale�) unless otherwise stated herein. No part of this content may be reproduced in any form, or referred to in any other publication, without the express consent of Grayscale.
10·CNeutral
H
Hanami8/20news
Zcash: Financial Privacy in the Age of AI
Key Takeaways
New technologies can upgrade the financial system but may also reduce financial privacy. This dynamic emerged with the digitization of banking and again with the rise of the internet. Today, the expansion of AI may create similar challenges.
Zcash @Zcash is a decentralized digital currency much like Bitcoin, but with built-in privacy features. In Zcash private transactions, the sender, receiver, and amount are all obscured, allowing Zcash to function like physical cash (which was Bitcoin’s original vision). In a world of AI-powered financial surveillance, these features could become essential.
Zcash is nearly 10 years old but may be entering a new chapter. Use of its shielding technology is increasing, and new capital is entering the ecosystem to support wallet development and Zcash mining.
Zcash’s ZEC token is valued at just ~$8 billion, or 0.6% of the Currencies Crypto Sector. If it were to capture just 5% of this market segment, its value would be ~9x greater. Zcash is a lower market cap crypto asset and should be considered higher risk.
Privacy is not a niche feature of money; it’s part of what makes money work. People don’t want every transaction, balance, and financial relationship exposed. Cash has this property. And historically, intermediated (bank-based) financial systems have preserved a level of practical confidentiality. They limit third-party access to financial records, though information can still be exposed. Even law enforcement generally requires a court order to access personal financial data.
Technological and social change can put pressure on these systems, often triggering public debate about financial privacy and new methods to protect it. In the 1970s, the digitization of financial records and the Bank Secrecy Act brought more attention to financial privacy (Exhibit 1). Similarly, in the late 1990s and early 2000s, the expansion of the internet and online banking, along with the Patriot Act, led to another wave of focus on financial privacy—and eventually broader use of encryption, 2FA, and related tools.
Exhibit 1: Technological change can bring focus to financial privacy
Today, Grayscale Research believes we are on the cusp of a third wave of widespread public attention on financial privacy, driven by the growth in stablecoins and other blockchain-based applications, as well as artificial intelligence (AI), which may offer new methods of financial surveillance. Based on higher-frequency indicators, public attention on these issues already seems to be increasing (Exhibit 2).
Exhibit 2: Recent increase in focus on financial privacy
This is where Zcash comes in. Zcash is a decentralized digital currency much like Bitcoin, but with sophisticated privacy features. Bitcoin made digital scarcity credible, but not digital privacy. Its transparency supports auditability, but limits money-like use cases where confidentiality matters. If cryptocurrency users eventually come to value private digital money as distinct from transparent digital assets, Zcash could capture part of that demand. At roughly 0.4% of total cryptocurrency market cap, ZEC appears small, in our view, relative to the quality of its technology and the potential addressable market of privacy-preserving digital money.
Privacy as a Product Category
On public blockchains, balances, counterparties, and transaction histories can remain visible indefinitely. Even when users transact through pseudonymous addresses, activity often becomes linkable over time—through exchanges, counterparties, wallet behavior, and blockchain analytics. In practice, these systems often reveal more information than users expect.
However, most users expect some level of financial privacy. Individuals may not want balances or spending history exposed by default. Businesses may want to keep suppliers, payroll, treasury flows, and customer relationships confidential. Institutions may not want wallet structures or transaction patterns mapped in real time. Privacy demand is therefore not limited to users seeking full anonymity. In many cases, it reflects ordinary preferences around confidentiality in economic life.
Yet, within crypto, the privacy category’s main issue is that it has often resulted in weaker distribution. Stronger privacy protections can create more friction in exchange support, wallet integration, and market access. As a result, privacy in crypto has not just been a technical feature; it has also involved a commercial tradeoff.
Approaches to Onchain Privacy
Not all crypto privacy systems hide the same information. Some mainly make transaction flows harder to trace on an otherwise transparent ledger. Others hide specific transaction fields more directly.
Mixing-based privacy on transparent chains such as Bitcoin CoinJoin, Bitcoin Cash CashFusion, and Dash PrivateSend try to make transaction flows harder to trace, but the underlying ledger remains visible.
Confidential transaction systems such as Litecoin MWEB can hide amounts and add structural privacy, while still differing from fully shielded systems.
Default-private systems such as Monero use built-in privacy mechanisms to obscure key transaction details by default, rather than asking users to opt in.
Shielded systems such as Zcash’s validate transfers without publicly revealing sender, recipient, or amount.
Within this privacy space, Zcash stands out as an optionally private base-layer system whose shielded transfers can hide sender, receiver, and amount at the transaction layer. These features place Zcash in a different category from privacy overlays on transparent blockchains (Exhibit 3).
Exhibit 3: Transaction privacy varies by cryptocurrency
What Zcash Is and How It Works
Zcash is a public blockchain for peer-to-peer value transfer with a fixed 21 million coin supply and a Proof of Work security model. In many respects, it resembles Bitcoin. The main difference is that Zcash gives users the option to shield transaction data rather than publishing all transaction details by default.
The network supports two transaction types.
Transparent transactions are publicly visible onchain, similar to Bitcoin-style transfers.
Shielded transactions verify that a transaction is valid without publicly revealing the sender, recipient, or amount.
In effect, Zcash allows verification without full public disclosure.
In Zcash, the privacy mechanism is a shielded pool validated with zero-knowledge proofs, which can reduce onchain disclosure substantially when users remain within shielded pools. Zero-knowledge proofs are a type of advanced cryptography that can be used for both privacy and scaling in a blockchain context.
Zcash also supports selective disclosure through viewing keys, allowing users to share visibility into shielded activity with specific third parties when needed. This matters because shielded transactions are not just simply “opaque”; they are private from the public while still allowing user-controlled disclosure when needed.
The Zcash Story: Old Protocol, New Chapter
Zcash launched in 2016. This makes it old by crypto standards and easy to misread as a holdover from an earlier market cycle. But from the beginning, it was more ambitious than a simple focus on privacy suggests. The project grew out of a serious effort to bring zero-knowledge cryptography into a digital cash system. Its core idea was unique at the time: combine a Bitcoin-like monetary structure with the ability to transact privately on a public blockchain. That made Zcash one of the earliest attempts to build private digital cash as a base-layer asset, rather than as a wallet feature or privacy layer added after the fact.
That ambition, however, ran ahead of the product infrastructure around it at the time. For much of its history, shielded functionality was harder to use than the transparent alternative because early shielded transactions were computationally demanding. Several major upgrades gradually reduced those constraints:
Sapling (2018): made shielded usage materially more practical by sharply reducing proving time and memory requirements relative to Sprout4. Zcash’s own materials say Sapling payments could be constructed in as little as a few seconds and with only 40 megabytes of memory, a change explicitly tied to improving adoption.
Orchard / NU5 (2022): modernized the cryptographic foundation by introducing the Orchard shielded protocol and Halo 2, removing the trusted-setup requirement for the new shielded pool.
Ironwood (2026): restored supply integrity by introducing a new shielded pool built on the corrected Orchard circuit, preventing any hypothetical counterfeit ZEC from expanding the circulating supply.
Unified Addresses / wallet improvements: helped reduce friction in wallet flows by making it easier for users to interact with both transparent and shielded receivers without relying on multiple address formats.
The challenge, in other words, was not just building private transfers. It was making them easier to use at scale. That long effort to reduce technical friction may finally be paying off.
One important piece of that story is the wallet layer. @zodl_co, formerly known as Zashi, is increasingly positioned not just as a place to hold ZEC, but as a shielded-first interface for using it. Current product materials highlight built-in swaps powered by Near Intents, CrossPay functionality that lets users send shielded ZEC while the recipient receives another asset, and support for cold storage of shielded balances.
These features matter because they reduce the number of steps where users would otherwise leave the shielded context to fund a wallet, make a payment, or move into another asset.
For Zcash, that’s important: the main bottleneck has historically not been whether private transfers were possible, but whether users could remain shielded through the full transaction flow.
There are also signs of broader infrastructure support around the network itself. Foundry, one of the largest cryptocurrency mining pools, announced plans to launch a US-based institutional Zcash mining pool in April 2026, aimed at institutional and public-company miners. That does not solve shielded wallet adoption directly, but it does suggest a different kind of ecosystem maturation: more industrial-grade infrastructure, more operational support, and potentially more institutional legibility around the network.
Finally, policy changes around fees and mempool behavior matter at the margin because they can make spam-style congestion less effective and improve network usability under load. That is not the core of the Zcash thesis, but it does fit the broader pattern: the network is being shaped not only for privacy in theory, but for more reliable use in practice.
The next phase of the Zcash story is likely to depend on whether the protocol can move from “privacy works” to “privacy scales.” The key development trends to watch are:
Tachyon, the clearest scaling proposal on the horizon.
At a high level, Tachyon aims to reduce some of the cost and synchronization burdens that have historically made shielded usage harder to support at scale.
Crosslink, an upgrade more about finality and network reliability than privacy itself.
The idea is to strengthen settlement assurances without replacing Zcash’s existing Proof of Work structure altogether. For the investment case, the relevance is that Zcash’s future is not only about privacy features in isolation; it also hinges on whether the broader network becomes more robust and easier to trust operationally.
Shielded assets, unlock beyond native ZEC.
If revived, they could extend Zcash’s privacy model beyond ZEC itself and make shielded transfer functionality relevant to a wider set of assets or applications.
What Is ZEC Worth?
As of July 2026, the market is assigning very little value to privacy in crypto.
Grayscale categorizes all the crypto assets with a “digital currency” use case in our proprietary Currencies Crypto Sector. Today, this crypto market segment includes 15 assets with a combined market capitalization of $1.4 trillion. Bitcoin is king of the category, with a market share of about 90%. Zcash’s ZEC token is valued at just ~$8 billion today, or 0.6% of the Currencies Crypto Sector (Exhibit 4). If it were to capture just 5% of this market segment, its value would be 9x greater.
Exhibit 4: ZEC’s share of the Currencies market cap is small but growing
ZEC’s current valuation appears to reflect an assumption that privacy will remain peripheral: useful to some, but too narrow to command a meaningful premium. But if privacy becomes more valuable in a world of greater surveillance, tighter compliance rails, and more visible financial censorship, then a 0.6% market share may look less like a fair equilibrium and more like a market that has not fully priced the asset's optionality.
The bullish version of the Zcash thesis is that the market is currently valuing ZEC as if privacy demand will stay contained, even though the chain already shows evidence that privacy usage is real. As of July 20, 2026, shielded transactions made up roughly 90% of Zcash transaction count, and shielded supply stood at about 4.2 million ZEC, or roughly 25% of current circulating supply (Exhibit 5). That suggests privacy is not merely a narrative attached to the asset, but an observable on-chain behavior.
Exhibit 5: ZEC’s shielded supply share has grown to an all time high
In other words, Zcash does not need explosive growth today to justify potential upside; it needs the market to change its view about what privacy is worth tomorrow. If investors eventually decide that privacy should command even a modest premium inside crypto's monetary stack, ZEC's current valuation could prove conservative, because the market is still treating privacy as more an afterthought than a core property of money.
Risks and Other Considerations
Zcash’s investment case is not just a question of privacy demand. It also depends on how investors think about legacy cryptographic assumptions, future security risks, and the network’s ability to keep shipping major upgrades without breaking coordination.
Regulatory risk
Zcash’s regulatory risk is not simply that it offers privacy, but that regulators and service providers may disagree on whether its selective disclosure tools are sufficient for compliance. Zcash’s viewing-key model gives a clearer compliance story than systems with no disclosure mechanism at all: a Full Viewing Key/Unified Full Viewing Key can be shared with a chosen party to reveal incoming activity and, for standard outgoing shielded payments, recipient address, amount, and memo. But that disclosure is still permissioned rather than default public traceability, while global AML/CFT standards for Virtual Asset Service Providers continue to emphasize customer due diligence, recordkeeping, suspicious activity reporting, and the Travel Rule, and Markets in Crypto Assets (MiCA) adds authorization, disclosure, and supervisory requirements in the EU. For an investor, the right framing is that viewing keys likely reduce—but do not eliminate—the risk that exchanges, custodians, or regulators treat shielded activity as operationally or politically harder to support.
Trusted-setup legacy pools
One historical concern around Zcash has been the trusted setup used for the older Sprout and Sapling shielded protocols. NU5 introduced Orchard and the Halo proving system, which removed the trusted-setup requirement for the Orchard shielded pool. Separately, a soundness vulnerability was discovered in the original Orchard circuit in 2026 that could theoretically have enabled undetectable counterfeiting within the Orchard pool. A network upgrade (NU6.2) corrected the circuit, while Ironwood (NU6.3) introduces a separate shielded pool. By preventing new value from entering the legacy Orchard pool and requiring funds leaving it to pass through Zcash’s turnstile accounting mechanism, Ironwood is designed to restore users’ ability to independently verify the integrity of Zcash’s circulating supply. This materially improves the forward-looking cryptographic story. The residual issue is that older pools still exist, and legacy funds can remain in them. In practice, this is best understood as a diminishing legacy risk rather than the defining risk of the current shielded design, but it is still relevant when comparing older and newer parts of the protocol.
Quantum risk
Quantum computing is a long-horizon risk for most public blockchains, not a Zcash-specific problem. At a high level, transparent and shielded funds may face different exposure paths depending on which cryptographic components are visible onchain and when spend-authorizing keys are revealed. Electric Coin Company, a company historically associated with building and launching Zcash, had recent roadmap work explicitly including quantum resilience as an area of focus6, which suggests the issue is being treated as a real design consideration rather than a purely theoretical one. For an investment note, the most reasonable framing is that quantum risk matters, but likely on a longer time horizon than the near-term usability, adoption, and market-structure questions that dominate the Zcash thesis today.
Execution risk
Zcash has historically relied on major protocol upgrades to improve usability, security, and scalability. That creates execution risk in two forms. First, the roadmap itself is ambitious: proposals such as Tachyon, Crosslink, and related changes are meaningful technical efforts rather than simple parameter tweaks. Second, network upgrades require coordination across protocol engineers, wallet developers, infrastructure providers, and the broader ecosystem. Zcash’s upgrade process is structured and well documented through Zcash Improvement Proposals, which is a strength, but it also means the investment case depends in part on continued coordination and implementation quality.
Conclusion
Privacy has always been part of money’s usefulness, even if digital systems have made that easier to forget. Zcash was built around a limitation of transparent blockchains: they make transactions easy to verify, but difficult to keep confidential. After years of technical work, it now stands as one of the clearest attempts to close that gap. Whether private digital money becomes a major category or remains a niche one is still uncertain. But at current valuations, the market appears to be assigning little probability to privacy becoming materially more important from here. For investors, that may be the opportunity: not a bet that Zcash replaces Bitcoin, but that the value of private digital money has yet to be fully recognized.
Investments in digital assets are speculative investments that involve high degrees of risk, including a partial or total loss of invested funds. Investments in digital assets are not suitable for any investor that cannot afford loss of the entire investment.
This information should not be relied upon as research, legal, tax or investment advice, or a recommendation regarding any products, strategies, or any investment in particular. This material is strictly for illustrative, educational, or informational purposes and is subject to change. This content does not constitute an offer to sell or the solicitation of an offer to sell or buy any security in any jurisdiction where such an offer or solicitation would be illegal. There is not enough information contained in this content to make an investment decision and any information contained herein should not be used as a basis for this purpose.
This content does not constitute a recommendation or take into account the particular investment objectives, financial situations, or needs of investors. Grayscale Operating, LLC and its affiliates, directors, officers, and employees may have interests, financial or otherwise, in the digital assets and/or securities discussed.
Certain of the statements contained herein may be statements of future expectations and other forward-looking statements that are based on Grayscale’s views and assumptions and involve known and unknown risks and uncertainties that could cause actual results, performance, or events to differ materially from those expressed or implied in such statements. In addition to statements that are forward-looking by reason of context, the words “may, will, should, could, can, expects, plans, intends, anticipates, believes, estimates, predicts, potential, projected, or continue” and similar expressions identify forward-looking statements. Grayscale assumes no obligation to update any forward-looking statements contained herein and you should not place undue reliance on such statements, which speak only as of the date hereof. Although Grayscale has taken reasonable care to ensure that the information contained herein is accurate, no representation or warranty (including liability towards third parties), expressed or implied, is made by Grayscale as to its accuracy, reliability, or completeness. You should not make any investment decisions based on these estimates and forward-looking statements.
There is no guarantee that the market conditions during the past period will be present in the future. Rather, it is most likely that the future market conditions will differ significantly from those of this past period, which could have a materially adverse impact on future returns. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN. PAST PERFORMANCE IS NOT INDICATIVE OF FUTURE RESULTS. We selected the timeframe for our analysis because we believe it broadly constitutes the most complete historical dataset for the digital assets that we have chosen to analyze.
All content is original and has been researched and produced by Grayscale Investments Sponsors, LLC (“Grayscale”) unless otherwise stated herein. No part of this content may be reproduced in any form, or referred to in any other publication, without the express consent of Grayscale.
10·CNeutral
m
meme8/19meme
quote: The AI race is moving down the stack.
Models → Compute → Energy.
The next layer to be rebuilt?
Capital.
When AI can reason and act autonomously, capital infrastructure can no longer be designed only for humans.
Machine-native capital markets are coming. | BREAKING: Nvidia, $NVDA, has agreed to provide a more than $100 billion backstop for a massive new OpenAI data center in Ohio, per FT.
Details include:
1. Nvidia will provide credit support for the "land, power and shell" for the facility capped at $105 billion
2. The data center is being developed alongside a SoftBank-led energy companies
3. Nvidia will also invest $1.5 billion into SB Energy, an energy company founded as part of the SoftBank Group in 2019
4. OpenAI plans to lease as much as 8 gigawatts of AI computing power at the data center in Pike County, Ohio, which will take until 2032 to complete
This will mark one of the largest AI infrastructure deals yet.
80·ALong
n
news8/18news
SanDisk Is Up 3,585%: Traders think it has more to run
The best-performing large cap in America over the past eighteen months is not an AI lab, a GPU designer, or a datacenter REIT. It is SanDisk, the company that made the memory card in your old camera. It listed at $48.60 on February 24, 2025, spun off from Western Digital to broad indifference. On Friday it closed at $1,790.82.
That is 3,585% in eighteen months, 49% of it in the last five sessions, and at roughly $278 billion of market cap it is now a top-30 US company.
Note: this is a stub of the full article on Hypercall Insights. Because of X platform limitations, the dropdowns, interactive components, and live data tools are not embedded here- read the full piece (with everything that did not fit) at https://insights.hypercall.xyz/sndk-supercycle-pricing-2026-08-17
Why is it performing so well? And what can the options market tell us about SanDisk's prospects? Let's start with the thesis.
AI inference is a storage problem
Everyone knows AI runs on GPUs and HBM. That story is three years old and fully priced. The newer, less-priced story is what happens one tier down.
A one-minute primer, for anyone who needs it. DRAM is working memory: each bit is a tiny capacitor that leaks, so it needs constant refresh and loses everything when power cuts, reads in nanoseconds, and costs real money per gigabyte. HBM is not a different silicon technology, it is DRAM packaged differently: dies thinned, stacked 8 to 12 high, mounted millimeters from the GPU with thousands of parallel connections. Same bits, vastly wider pipe, several times the cost. NAND flash is storage: bits are trapped charge that stays put with the power off, cells stack in 3D by the hundreds of layers, so it costs roughly a hundredth of HBM per byte. The price is speed, microseconds instead of nanoseconds, and cells wear out with heavy writing. Every SSD, memory card, and phone is NAND. This is what SanDisk makes. (A good visual explainer of how flash works: youtube.com/watch?v=dZcszUj5szA)
Here is the hierarchy every AI system lives inside. Fast memory is tiny and obscenely expensive. Big storage is cheap and far too slow. Every byte an AI system touches has to find a home somewhere on this ladder:
HBM gets the headlines because it is bolted to the GPU. But HBM is sized for compute, not for state, and modern AI systems generate state in absurd quantities. Count what actually has to live somewhere:
Two of these deserve a closer look, because they are the ones growing fastest.
Weights: the four-orders-of-magnitude decade
Model parameters map almost one-to-one to bytes. When models were 1.5 billion parameters, weights were a rounding error. They are not a rounding error anymore:
The one force pushing the other way arrived in 2024: fp8 and int4 quantization plus sparse mixture-of-experts cut the bytes actually served per parameter by 2 to 4x. The totals exploded anyway. A frontier-class model is still measured in terabytes, and no serving fleet holds one copy. It holds a copy per node, per region, per fine-tune, per experiment. The weights themselves have become a distribution problem that only flash is fast enough and cheap enough to solve.
KV cache: the state that ate the datacenter
The really fun one is the KV cache, the keys and values every transformer layer stores for every token of context, the working memory of a conversation. Every token of context a transformer holds costs memory for the rest of the conversation, roughly half a megabyte per token for a 405B-class dense model. That sounds harmless until you multiply it by agents running million-token contexts and thousands of concurrent sessions. (The article has an interactive calculator here; the snapshot below shows one setting.)
Play with that for thirty seconds and you understand the entire trade. A handful of long-context sessions overwhelms the HBM on a node that costs as much as a house. The industry's answer is not "buy more HBM," because there is no more HBM. The answer is to tier the state: park cold conversations on flash, reload them when the user comes back. SanDisk management now sizes this single workload, KV cache offload, at 75 to 100 exabytes of potential 2027 demand. For scale, that is roughly a third of the entire industry's annual output, from a workload that barely existed two years ago.
Add the embeddings behind every RAG system, the checkpoints behind every training run, and the hundred-petabyte corpora that dataloaders hammer at random, and the conclusion writes itself: the marginal byte of AI infrastructure is increasingly a flash byte. Storage stopped being the boring aisle of the datacenter sometime in 2025. The market took a while to notice.
The shortage
Demand was only half of it. The 2022-23 memory bust was bad enough that every NAND maker cut capex and idled lines, so when AI demand arrived there was nothing spare to sell:
Flash fabs are not light switches. Contract NAND pricing is now forecast up 75 to 100%, and since a wafer costs the same to make at any selling price, most of that increase falls straight to margin. SanDisk says about two thirds of its June-quarter sequential growth came from price, not volume.
Implied vol across the storage complex shows how precisely the market has sorted this out:
Micron makes DRAM and HBM alongside NAND. Western Digital kept the hard drives. Seagate is drives with a NAND garnish. SanDisk is the only large-cap pure play on flash pricing, and it carries a 13 to 20 vol point premium over all of them, at every tenor, out to 2028. The purer the exposure, the wider the distribution.
Purity is priced. SNDK at 91% implied vol against Micron's 71% is not the market calling SanDisk a worse company. It is the market calling it the least-hedged bet on the same question. If NAND pricing holds, SNDK earns the most per dollar of market cap. If it cracks, there is no DRAM division to hide behind.
Datacenter took over in four quarters
That premium is earned. Revenue by end market, every quarter since the spin-off:
Datacenter went from $213 million to $2,977 million in four quarters and is now a third of revenue, growing 103% sequentially. Consumer, the SD cards the company was named after, shrank 5%. Even the boring Edge segment, flash sold into PCs and phones, quadrupled on pricing alone.
Margins are where it stops looking like a memory company at all:
84.6%. Gross margin, June quarter. A company selling a commodity into a spot market cannot print that. A company selling allocation of a scarce resource under contract can. For calibration: TSMC runs high-50s, Nvidia mid-70s, and SanDisk itself printed 26.2% a year ago.
The contracts
Those contracts got a dollar figure on August 13, at SanDisk's Investor Day: $93.9 billion of contracted business under its "New Business Model" agreements. Multi-year hyperscaler deals, quantities detailed by month, fixed price floors, financial guarantees. Ten signed, five since April. The stock rose 49% that week.
Management's fiscal 2028-2030 targets off this base: mid-to-high-teens revenue growth, gross margins around 80%, free cash flow margins near 50%, a buyback authorization now at $15.5 billion. Toll-road economics, if you believe them.
The bull case for the contracts is that they break the destocking spiral that made every previous memory downturn worse. Customers with take-or-pay floors do not run down inventory on a whim, and a hyperscaler signing four years of volume is showing you its own internal demand forecast.
The bear case is polysilicon. After 2011, ten-year take-or-pay contracts got renegotiated, litigated, or simply walked once spot fell far enough below contract. Long-term agreements dampen cycles; they do not repeal them. What $93.9 billion really buys is a change of question, from "will demand hold" to "will these prices hold when supply arrives." A better fight. Still a fight.
HBF: flash on the GPU package
There is one more leg, further out, and it needs a paragraph of packaging to make sense.
HBM is fast for a packaging reason, not a silicon one. Take ordinary DRAM dies, thin them, stack eight or twelve high, drill thousands of vertical connections through the stack, and mount the whole thing millimeters from the GPU on a shared slab of silicon. Width does the work: the stack talks to the GPU over thousands of wires at once, which is how you get terabytes per second out of memory that is not individually fast. The cost is capacity. DRAM cells are large, stacks have height and heat limits, and you end up with 36 to 48 GB per stack of the most expensive memory ever mass-produced.
High Bandwidth Flash is the same packaging trick with NAND dies in the stack instead of DRAM. NAND stores an order of magnitude more bits per die, so the published spec lands at 512 GB per stack at 1.6 TB/s: HBM-class bandwidth, more than ten times HBM capacity.
Flash has two real handicaps as memory. Reads take microseconds instead of nanoseconds, and cells wear out if you write them constantly. Both handicaps miss the inference workload almost entirely. Weights are written once and read billions of times. KV cache reloads are big sequential streams, the one access pattern where flash latency hides behind bandwidth. HBF would be useless as general-purpose memory, and it is not aimed at general-purpose memory. It is aimed at the workload from the first section of this piece.
If it ships, eight stacks put about 4 TB next to one accelerator, room for an entire frontier model's weights on the package, where today's HBM holds a fraction of them. SanDisk co-authored the spec with SK hynix and published it through the Open Compute Project in early August. First samples are due in the second half of 2026, first inference devices in early 2027.
None of it is in current earnings, and none of it should be modeled as revenue. It is a call option stapled to the stock: right partner, open standard, right workload. Paying something for it is reasonable.
Whose news moves whom
SanDisk, Micron, Western Digital, and Seagate now trade as one macro bet. SNDK and MU daily returns have correlated at 0.85 over the past 60 days. The options market prices them together too, but not identically, and the differences are where the information is.
The level difference is the purity premium from earlier: SNDK at 91%, the rest 13 to 20 points below, at every tenor. Same question, different leverage to the answer.
The more interesting difference is what each surface says about whose news matters. Micron reports on roughly September 22. None of the other three companies has any event that week. Their forward curves know anyway:
Micron's own report week prices 12 points over its neighbors, which is what an earnings week looks like. But Seagate, with nothing on its calendar, carries the same 12-point bump, and SanDisk carries five and a half. Only Western Digital, the one name with no NAND on the income statement, sleeps through it. In a shortage, the first supplier to disclose contract pricing moves everyone who sells flash.
Run the test backwards and it fails. SanDisk reports November 5, and its own forward for that window runs 96.7%, eleven points over base. Micron's forward through the same window: 71.2%, between a 70.0% stretch before and 69.0% after. SanDisk's biggest day of the year does not exist on Micron's surface. The market has decided information flows one way down this supply chain, from the diversified bellwether to the leveraged pure play, never back.
The leverage is quantifiable. Total variance across SNDK's expiries fits base-plus-events with about one vol point of error: 85.8% base, plus a 17.7% move for each of the next seven earnings reports. The same fit on Micron gives 70.6% and 4.2%. A Micron print is a data point. A SanDisk print is a referendum, priced at roughly double anything SanDisk has actually delivered:
We do not think that repricing is crazy. Under the NBM model, earnings day became the day the contracted book gets marked in public: new signings, new floors, new customers. November 5 is the first print against both the $10.3 to $10.8 billion guide and the Investor Day targets.
The far end of the curve makes the same point on a longer clock:
The top of the entire curve, 99.4%, sits in the first half of 2028, and stays at 91.3% even after stripping out both earnings reports inside that window. That is not an earnings hump. It lines up with the industry's own supply calendar:
Until late 2027 the shortage is unfalsifiable: no quarter can prove the bears right while the fabs that could oversupply the market do not exist. The first real test of the $93.9 billion book against new capacity comes in 2028, and the surface has parked its maximum uncertainty exactly there.
The market prices the debate, not the answer. The last question in every SanDisk bull-bear argument is whether AI created a structural NAND supercycle or a spectacular but temporary squeeze. The options market's answer: still an open question in 2028, litigated in 17.7% increments every quarter until then, with the widest distribution of outcomes parked exactly where new supply can first arrive.
In dollars, here is the whole piece on one time axis: the realized path in, the implied distribution out:
A range of $509 to $6,361 by mid-2028 is the market holding both endings live: the commodity cycle in a party hat, and the structurally-80%-margin infrastructure company. The 25-delta risk reversal is inverted, calls over puts by 3 to 4 points at every expiry, so of the two tails, the market pays up for the melt-up. Meanwhile the equity trades near 10 times annualized guided earnings, a multiple that says "temporary." The stock and its own options are having an argument. We would rather own the argument than either side of it.
The bottom line
AI made storage a first-order input. That demand hit the one commodity whose supply had spent three years shrinking, at the purest-play supplier, which converted the squeeze into four years of contracts and margins no memory company has printed before. Nobody finds out whether it lasts until supply returns in 2028. Until then the path itself is priced: seven referendums at 17.7% each, a checkpoint every time Micron speaks, and a final exam in the first half of 2028.
SNDK is live on Hypercall. Everything in this piece, the term structure, the event pricing, the 2028 forwards, is a market you can trade, on-chain: app.hypercall.xyz/asset/sndk. The full interactive version, with the KV-cache calculator and the complete methodology, is at insights.hypercall.xyz/sndk-supercycle-pricing-2026-08-17.
Data, methods, sources, and caveats