Market Brief: Friday, November 6, 2026

Today's lesson

How to compare model quality vs cost

Learn how to compare AI model benchmark performance with token pricing, latency, throughput, and cost per useful result. Why it matters today: Today's lesson connects the Model Benchmarks & AI Compute Economics track to the compute market readers are watching.

Term of the Day

Scale-up vs scale-out

Scale-up connects accelerators tightly within one system or rack. Scale-out links multiple systems through a larger network. Large AI clusters often use both. GPUs may exchange data over NVLink inside a rack and use another network to communicate with neighboring racks.

H100 IndexiWhat it costs to rent one H100 GPU for an hour on demand — the market’s anchor price.

Value

$5.42 / H100-hr

Last Price Change

Nov 9, 2026

Spot Compute PriceiThe cheaper, interruptible GPU price — a read on whether spare capacity is available.

Value

$2.46 / H100e-hourH100e-hour means an H100-equivalent GPU hour, a normalized unit used to compare different GPUs and providers.

Last Price Change

May 27, 2026

Model Serving CostiRoughly what it costs to run a model and generate its output at scale, per million tokens.

Value

$2.78 / 1M tokens

Last Price Change

Aug 8, 2026

GPU AvailabilityiHow easy or hard it is to actually get GPU capacity right now, scored 0–100.

Value

83 / 100 score

Last change

Sep 17, 2026

Nov 6, 2026

ING Bank looks to sell $500m in data center loans at once, including at Stargate site

The project described in “ING Bank looks to sell $500m in data center loans at once, including at Stargate site” adds prospective compute capacity, but permitting, power procurement, construction, and deployment timing prevent it from increasing immediately available GPU supply.

Nov 6, 2026

DTE Energy ups five-year utility investment plan to $44bn, cites data center demand

The project described in “DTE Energy ups five-year utility investment plan to $44bn, cites data center demand” adds prospective compute capacity, but permitting, power procurement, construction, and deployment timing prevent it from increasing immediately available GPU supply.

Nov 6, 2026

Ongoing labor dispute sees Micron pause construction at its $4.5bn memory fab in Taoyuan, Taiwan – report

The source for “Ongoing labor dispute sees Micron pause construction at its $4.5bn memory fab in Taoyuan, Taiwan – report” does not yet establish a supported change in compute supply, demand, price, availability, power delivery, financing, or deployment timing.

Nov 9, 2026

Building the next generation of multimodal AI experiences with fal in Microsoft Foundry

The source for “Building the next generation of multimodal AI experiences with fal in Microsoft Foundry” does not yet establish a supported change in compute supply, demand, price, availability, power delivery, financing, or deployment timing.

Nov 9, 2026

PJM Winter Outlook: Power Supplies Available to Serve Expected Cold Weather Conditions

The source for “PJM Winter Outlook: Power Supplies Available to Serve Expected Cold Weather Conditions” does not yet establish a supported change in compute supply, demand, price, availability, power delivery, financing, or deployment timing.

H100 GPU Pricing Report: September 2026

Canonical CT-H100 median, reviewed H100 provider rows, freshness dates, and composition caveats in one citable report.

Previous issue: Market Brief: Thursday, November 5, 2026

Sempra expects 20% increase to Oncor’s capex plan, driven by load growth

Next issue: Market Brief: Monday, November 9, 2026

Sempra expects 20% increase to Oncor’s capex plan, driven by load growth