AI compute market signals
NewsletterContact

Daily Brief

ComputeTape Daily Brief

AI compute supply, demand, pricing, and infrastructure signals.

A concise pre-market read on what changed in AI compute and what to watch next.

The Tape

Today’s Tape

  • AMD and Meta Announce Expanded Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs.
  • AMD and OpenAI Announce Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs.
  • FlashAttention-4 gives the NVIDIA Blackwell platform its most optimized attention kernel yet.

Indexes

ComputeTape Index Snapshot

$8.41 / H100e-hrCT-COST

No prior-day comparison. Enterprise AI cloud cost.

$2.73 / spot H100e-hrCT-SPOT

No prior-day comparison. Interruptible compute cost.

$6.60 / market H100e-hrCT-MKT

No prior-day comparison. Observed market-dollar H100e rate.

$70,866 / training runCT-TRAIN

No prior-day comparison. Standard 7B / 1T-token training run.

$3.40 / 1M tokensCT-INFER

No prior-day comparison. Standard 70B-class inference workload.

70.0 / 100 availabilityCT-GPUA

No prior-day comparison. Practical GPU availability.

Latest Brief

Top market signals

Start here. These are the highest-impact source-backed items from the Daily Brief.

AMD and Meta Announce Expanded Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs

AMD and Meta Announce Expanded Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs. Why it matters: this may affect gpu supply, demand, capacity, bottlenecks, or financing.

Source: AMD Investor Relations

AMD and OpenAI Announce Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs

AMD and OpenAI Announce Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs. Why it matters: this may affect gpu supply, demand, capacity, bottlenecks, or financing.

Source: AMD Investor Relations

FlashAttention-4 gives the NVIDIA Blackwell platform its most optimized attention kernel yet

FlashAttention-4 gives the NVIDIA Blackwell platform its most optimized attention kernel yet. Why it matters: this may affect gpu supply, demand, capacity, bottlenecks, or financing.

Source: Lambda Blog

Microsoft Databases and Microsoft Fabric: Your unified and AI-powered data estate

Microsoft Databases and Microsoft Fabric: Your unified and AI-powered data estate . Why it matters: this may affect power supply, demand, capacity, bottlenecks, or financing.

Source: Microsoft Azure Blog

NVIDIA and Corning Announce Long-Term Partnership to Strengthen US Manufacturing for AI Infrastructure

NVIDIA and Corning Announce Long-Term Partnership to Strengthen US Manufacturing for AI Infrastructure. Why it matters: this may affect gpu supply, demand, capacity, bottlenecks, or financing.

Source: NVIDIA Newsroom

NVIDIA and IREN Announce Strategic Partnership to Accelerate Deployment of up to 5 Gigawatts of AI Infrastructure

NVIDIA and IREN Announce Strategic Partnership to Accelerate Deployment of up to 5 Gigawatts of AI Infrastructure. Why it matters: this may affect gpu supply, demand, capacity, bottlenecks, or financing.

Source: NVIDIA Newsroom

OpenAI’s GPT-5.5 in Microsoft Foundry: Frontier intelligence on an enterprise ready platform

OpenAI’s GPT-5.5 in Microsoft Foundry: Frontier intelligence on an enterprise ready platform . Why it matters: this may affect cloud supply, demand, capacity, bottlenecks, or financing.

Source: Microsoft Azure Blog

Capacity Watch

Capacity Watch

  • Hermes Unlocks Self-Improving AI Agents, Powered by NVIDIA RTX PCs and DGX Spark (NVIDIA Blog).
  • NVIDIA Spectrum-X — the Open, AI-Native Ethernet Fabric — Sets the Standard for Gigascale AI, Now With MRC (NVIDIA Blog).
  • OpenAI’s New GPT-5.5 Powers Codex on NVIDIA Infrastructure — and NVIDIA Is Already Putting It to Work (NVIDIA Blog).

Power & Data Center Watch

Power & Data Center Watch

  • Powering the Next American Century: US Energy Secretary Chris Wright and NVIDIA’s Ian Buck on the Genesis Mission (NVIDIA Blog).
  • Microsoft Databases and Microsoft Fabric: Your unified and AI-powered data estate (Microsoft Azure Blog).
  • OpenAI’s GPT-5.5 in Microsoft Foundry: Frontier intelligence on an enterprise ready platform (Microsoft Azure Blog).

Capital Flow Watch

Capital Flow Watch

  • Release announcement on GitHub (Google Cloud Release Notes).
  • AMD Powers U.S. Sovereign AI Factory Supercomputers, Accelerating an Open American AI Stack (AMD Investor Relations).

Analyst Note

Why this matters

Today’s brief emphasizes the highest-scoring signals tied to AI compute cost, availability, demand, power, and capital flows.

  • Today’s brief emphasizes the highest-scoring signals tied to AI compute cost, availability, demand, power, and capital flows.
  • AMD and Meta Announce Expanded Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs affects CT-COST, CT-TRAIN, CT-INFER, CT-GPUA.
  • AMD and OpenAI Announce Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs affects CT-COST, CT-TRAIN, CT-INFER, CT-GPUA.

Watch next

What could move the next Daily Brief

  • Hermes Unlocks Self-Improving AI Agents, Powered by NVIDIA RTX PCs and DGX Spark (NVIDIA Blog).
  • NVIDIA Spectrum-X — the Open, AI-Native Ethernet Fabric — Sets the Standard for Gigascale AI, Now With MRC (NVIDIA Blog).
  • OpenAI’s New GPT-5.5 Powers Codex on NVIDIA Infrastructure — and NVIDIA Is Already Putting It to Work (NVIDIA Blog).
  • Powering the Next American Century: US Energy Secretary Chris Wright and NVIDIA’s Ian Buck on the Genesis Mission (NVIDIA Blog).
  • Microsoft Databases and Microsoft Fabric: Your unified and AI-powered data estate (Microsoft Azure Blog).

Recent Briefs

Last 10 Briefs

Monday, May 18, 2026

Daily Brief: 2026-05-18

AMD and Meta Announce Expanded Strategic Partnership to Deploy 6 Gigawatts of AMD GPUs affects CT-COST, CT-TRAIN, CT-INFER, CT-GPUA.

Open brief →