01
Reserve ownership
Own the scarce monetary asset before AI-infrastructure stress becomes obvious, rather than trying to acquire it only after liquidity and correlations have changed.
BTX converts surplus AI infrastructure into fixed-supply monetary security.
A 21-million-maximum, post-quantum monetary asset secured by deterministic work aligned with the accelerator, memory, power and cooling stack that carries the AI capital cycle.
Own AI for the expansion. Own BTX for the contraction.
The complete institutional structure combines three separately governed pieces: a BTX reserve, compatible infrastructure, and contractual switching, step-in or collateral-control rights. No one piece should be mistaken for the whole hedge.
01
Own the scarce monetary asset before AI-infrastructure stress becomes obvious, rather than trying to acquire it only after liquidity and correlations have changed.
02
Redirect qualified capacity when BTX netback exceeds the available commercial AI netback, subject to contracts, power, software and device qualification.
03
Give lenders, lessors or insurers an additional operating and recovery path after a trigger—when enforceable access, step-in, custody and proceeds rights already exist.
Compatible infrastructure serves the highest-value commercial AI workload.
Idle or underused qualified capacity becomes available for an alternative workload.
Competition and network security can rise as deterministic work joins proof of work.
Scheduled issuance and the 21 million maximum do not expand with capacity.
Security-production capacity is elastic. Monetary supply is inelastic.
AI infrastructure is financed as equipment, energy load, specialized real estate and credit at the same time. A derivative can settle correctly while the owner still faces a dark data hall, fixed power costs and hardware every other distressed seller is trying to place.
The correct role
BTX complements puts, credit protection, insurance, power hedges, compute forwards and liquidity reserves. It targets the residual utilization and recovery problem those instruments do not solve.
Demand can fall below underwriting assumptions; customers can renegotiate, fail or decline to renew.
Accelerators, interconnects, memory and facility design can lose economic value faster than scheduled depreciation.
Take-or-pay energy commitments and debt maturities can survive after the commercial workload weakens.
Many owners may need to sell or re-lease similar capacity at the same time.
The report expected Profile 2 and Coupled-V3 to become final. The current release instead schedules MatMul v4.7 Epoch A Profile 1 ExactReplay at mainnet height 185,000. The institutional thesis remains useful, but every technical claim must follow the shipped code.
Twenty-one million maximum units, open work-based issuance and holder-controlled transfer—without a corporate redemption promise.
Ordinary P2MR transfers use the NIST-standardized ML-DSA and SLH-DSA signature families in production.
Epoch A uses complete deterministic Profile 1 replay on qualified Metal or CUDA paths; CPU replay is diagnostic, not an automatic production fallback.
Below block 185,000 mainnet remains on MatMul v3. At the boundary it atomically enters v4.7 Epoch A. Profile 2, Coupled-V3 and succinct-proof authority remain disabled in v0.33.2.
The protocol creates the asset; markets and operating standards make it usable. BTX needs competing production, OTC and wholesale inventory, executable public discovery, qualified custody, documented financing rights and strategic end demand.
Allocators, insurers, treasuries, utilities and sovereigns hold long-duration BTX exposure.
Qualified custody, funds, indices, lending controls and policy-grade reporting.
Verified deposits, withdrawals, bids, asks, depth and trade reporting—not reference values alone.
Competing dealers consolidate production and carry timing, custody and counterparty risk.
Independent operators settle recurring supply under documented terms.
Permissionless proof of work creates supply without a corporate unlock or discretionary treasury.
Qualified accelerators, memory, power, cooling, networks and operator skill form the productive base.
Physical capacity ↔ issuance ↔ miner / OTC liquidity ↔ wholesale inventory ↔ public discovery ↔ custody and products ↔ reserve demand.
Explore market access →Compatible hardware is not enough. A credible switching option needs qualified devices, current signed software, power and network capacity, contractual permission, custody controls and a rehearsed operating plan before distress.
When machines leave mining they remain general-purpose infrastructure for open models, agents and numerical workloads. BTX does not place subjective AI tasks inside consensus.
Inventory exact accelerator, memory, driver and runtime combinations; do not assume any GPU qualifies.
Pin and verify v0.33.2 source or signed binaries, checksums and activation configuration.
Model marginal energy, cooling, interconnection, uptime and facility constraints.
Customer, lender, lessor, site and power agreements must expressly permit the operating state.
Separate reserve, treasury, node, attestation and recovery authority with tested succession.
Test qualification, fail-closed behavior, restart, payout and incident procedures before relying on the option.
Institutional positioning must map to validation rules, device qualification and active work construction. The correct claim is a height-versioned deterministic accelerator workload—not that every planned research component is live.
Fixed maximum supply with work-based issuance.
Production ordinary transfer is constrained to the post-quantum P2MR path.
Mainnet keeps the existing 512×512 M31 work function below height 185,000.
The boundary atomically activates Resident Curriculum Profile 1 ExactReplay.
Every claimed Epoch-A block requires the complete deterministic four-round Profile 1 replay.
Production is qualified around Metal or CUDA; Profile 2, Coupled-V3 and succinct-proof authority remain disabled.
Product labels do not create legal or operating rights. Each structure must separate the BTX asset, infrastructure, control rights, custody, valuation, triggers and waterfall—and underwrite each channel independently.
Hold a governed BTX allocation alongside AI equity, private-credit or infrastructure exposure.
Review the institutional thesisAdd tested step-in, equipment, power, software, wallet and proceeds rights to the recovery set.
Review credit requirementsPre-negotiate the ability to redirect qualified idle capacity when contracts and netback permit.
Open mining operationsPair AI-related exposure with a post-quantum reserve and a controlled domestic compute fleet.
Review the reserve frameworkPreserve an alternative utilization path without assuming guaranteed mining profit.
Expand the recovery set beyond liquidation or re-leasing alone.
Own the scarce monetary network that may receive released capacity.
The correct comparison is functional. Traditional tools remain better for precise triggers, named counterparties and defined horizons. BTX adds a reserve-and-recovery mechanism tied to the physical capacity creating the exposure.
These can pay against defined market or credit events. They do not create a new use for an idle accelerator fleet.
These can stabilize input or rental-price indices. They do not solve facility-specific vacancy, obsolescence or step-in execution.
These can preserve liquidity or diversify macro risk. They have no operating relationship with AI hardware.
The stronger and more defensible claim: BTX creates a new reserve-and-recovery mechanism that conventional products do not provide. It is not a guaranteed inverse instrument.
Track executable depth, dealer diversity, custody, qualified capacity, mining distribution, device readiness, source releases and contractual control—not only a model value or headline token price.
Claims to avoid
The complete proposition is a scarce reserve asset, a utilization backstop and a collateral-recovery option—implemented with qualified infrastructure, enforceable rights and independent risk controls.