Browse documentation
DocsMarkets & economics
Proposed market design

Spot capacity, monthly forwards and compute hedges

The investor brief proposes spot capacity for near-term delivery, monthly forwards or futures for planned requirements, and perpetual-style exposure to a compute-price index. These products have different delivery and risk obligations. The current interface stages examples; it does not operate a live matching or derivatives venue.

On this page

Three products with different obligations#

A buyer planning an actual workload should begin with delivery requirements. An index-linked position may offset some price exposure, but it does not ensure that the required hardware is available in a particular region or window. That distinction must remain visible in order tickets and documentation.

The brief uses the combined planning term forwards/futures. It does not settle a legal classification or fully specify clearing, transferability and delivery election. Those decisions require a formal product specification before a production market is offered.

Three products with different obligations
ProductIntended useKey boundary
Spot capacityNear-term compute deliveryConfirm hardware, scheduling and verification terms
Monthly forward / futurePlan demand or pre-sell committed supplyDefine expiry, marking, collateral and delivery or settlement obligations
Perpetual-style hedgeOngoing exposure to a compute-price indexFunding and margin are distinct from physical GPU access

A market needs more than a symbol#

The initial calendar is monthly, with daily mark-to-market and final settlement at expiry proposed for the dated products. A symbol is a shorthand for hardware, region and period. It does not replace the full terms, including eligible providers, quantity, verification, currency and failure procedures.

The production specification must settle open questions such as partial fills, minimum size, cancellation, expiry cutoffs, position limits and the relationship between a traded contract and an executable reservation. A UI dropdown for spot, forward or perp does not implement those rules.

Order book first, AMM fallback proposed#

A central limit order book is proposed for explicit limit prices and larger quotes. An automated market maker would provide an alternative quoting path when the book is thin. These are complementary matching concepts, not a guarantee that every order can execute at a predictable total cost.

A production router must define price limits, slippage, fees, sequencing and how quote validity is checked. It must also prevent a displayed AMM quote from implying delivery capacity that the provider layer has not committed. Neither a live order book nor an operational AMM exists in the inspected Node scaffold.

Treasury liquidity has a separate risk budget#

The brief names the NeuraMesh treasury as the proposed initial liquidity source. It also requires position limits, transparent inventory reporting, a separate risk budget and governance-approved withdrawal rules. Treasury-backed quotes should not be described as independent third-party demand.

Liquidity and capacity are different resources. An account can have enough collateral to trade while the underlying market lacks deliverable GPUs. A provider can have hardware while a thin market produces poor price discovery. Readiness requires testing both constraints.

Measure usable markets, not listing count#

The brief prioritizes bid-ask spread, depth, traded notional, open interest, buyer repeat rate, provider fill rate and delivered GPU-hours. These need precise definitions and a separation between simulated, treasury-generated and external activity.

Adding H100, H200, A100, L40S, RTX 4090, MI300X, B200 or multi-GPU markets is a proposed pilot direction subject to liquidity and approved supply. It is not a statement that any of these devices can currently be booked through NeuraMesh.

Common questions

Does a perpetual hedge give me GPU access?

No. Index exposure is not a physical reservation. Scheduling rights require explicit capacity and delivery terms.

Are the displayed market prices live?

No NeuraMesh market price exists. GET /api/markets publishes none. The workspace shows real third-party share prices for listed compute companies and real external GPU rates, plus a simulated GPU-hour history because no free feed publishes one. None of them is a NeuraMesh executable quote.

Sources & implementation status

Based on the NeuraMesh investor brief and the project sources below. Source review: . Design explanations are not a deployment audit, legal opinion or commitment to launch.

Read our source and status conventions. Suggest a correction.