Delivered Compute Report — #1

2026-09-10 · Compute Assay · spec v0.1

Three exchanges are launching cash-settled GPU price futures. Every one of them tells you what compute costs. None of them tells you what you'd actually receive. This report covers the second question: what capacity really exists, how it's wired, and whether it can do the job the price implies.

Surveyed: 20 providers, 133 offerings (public rate cards and docs only — everything cited).

Correction (2026-09-14). This report originally stated 16 providers / 70 offerings and a 6.2× H100-SXM spread. Both numbers were wrong, in two ways we found during a self-audit. First, a pipeline validation bug silently dropped an entire input file — RunPod, TensorDock, and Prime Intellect (16 offerings) were surveyed on Sep 10 but never entered the registry; the pipeline now fails loudly instead of shrinking quietly. Second, the 6.2× spread was anchored on Voltage Park's $1.99 — which we verified is a quote-gated "starting at" teaser (it no longer appears on their pricing tiers at all), not a bookable walk-up price. Teaser, floor, and illustrative prices are now excluded from every headline statistic and labeled in every table. The restated headline: $3.20–$12.29, a 3.8× walk-up spread. The registry improves by being wrong in public; every revision is checkable against the immutable snapshot archive.

Headline prices ($/GPU-hr, on-demand)

The same H100-SXM sells for $3.20 and for $12.29 this week — a 3.8× gap. A price ratio that size between sellers of "the same" good is not a spread; it is several different goods sharing a name. The table below is why.

The deliverability gap

We grade every offering on whether it can actually run the workload the price implies (spec §4: scale-up domain × scale-out fabric × contiguity):

48 of 133 surveyed offerings are verifiably training-class. 14% publish nothing about interconnect — capacity you cannot underwrite, price, or write a delivery contract against. That opacity is the market's missing half.

Cheapest verifiable training-class capacity (Grade A/B)

Residual watch: what the hardware itself is worth

Secondary-market observations (collected 2026-09-14) — the collateral question under every GPU-backed loan. Ranges span markets: auction clears run well below reseller asks, and that spread is the LTV argument.

Every observation carries its source URL in residual-values.json. Rental prices are the flow; this is the stock — when the two diverge, one of them is wrong.

This week's read

The number everyone will trade against next month is $2.53 — Silicon Data's H100 rental print, the index underneath the CME contracts listing October 5. Here is what that number is hiding: Silicon Data's own hyperscaler-tier H100 index sits near $7.48, and in this week's survey the same chip clears at $1.97 on the open marketplace, $3.20–$3.99 at neoclouds with a stated InfiniBand fabric, and $6.88–$12.29 at hyperscalers. That is not one commodity with a wide spread; it is at least three different goods sharing a ticker. The difference between them is not brand — it is whether anyone will state, in writing, what the network looks like.

Three details from this week's data that make the point:

1. Training-class capacity does not trade on the open market. Our live sweep of the largest GPU marketplace found zero 8-GPU H100-SXM hosts — the largest contiguous H100-SXM listing was 4 GPUs, B200 the same. Anything that can actually run a distributed training job is locked in reserved contracts, which is precisely the capacity a cash-settled future cannot reach.

2. Hyperscaler list prices are administrative, not market-clearing. Oracle prices H100 and H200 identically at $10.00/GPU-hr. Azure's H200 lists below its H100 ($10.60 vs $12.29). Google publishes no on-demand price for H200 or B200 at all — those chips are reservation-only. A hedger settling against a hyperscaler-tier reading is hedging a posted price, not a market.

3. 40% of surveyed capacity can't verifiably do the job its price implies. Of 70 offerings across 16 providers, only 10 meet the bar for large-scale training on published specs (Grade A), and 6 publish nothing about interconnect at all. The sellers who publish the most fabric detail — Azure, Oracle, Nebius, CoreWeave — are also the ones whose capacity you could actually write a delivery contract against. That correlation is the entire case for a deliverable spec.

One more thing worth watching: the physical layer is starting to move. Architect and Compute Desk announced ComputeConnect in July — an exchange-for-physical mechanism for converting compute futures into real capacity — and SF Compute is running a live spot/forward cluster market. An EFP network without a standardized definition of the deliverable is a settlement dispute waiting to happen. Somebody has to write the spec. That's what this report is.

Method

Prices and specs collected 2026-09-10 from public rate cards and official docs only; every row carries its source. Fields a provider does not publish are recorded as unknown, never estimated. Deliverability grades follow the Deliverable Compute Registry spec v0.1. Corrections welcome — the registry improves by being wrong in public.