Layer 2.5 · head to head
AWS (P5, P6, G6) vs Google Cloud (A3, A4)
Capacity Blocks against Dynamic Workload Scheduler — two answers to the same shortage.
| AWS (P5, P6, G6) | Google Cloud (A3, A4) | |
|---|---|---|
| Can you leave | Movable with effort. Expect to redo images, storage wiring and networking. | Movable with effort. Expect to redo images, storage wiring and networking. |
| Counterparty | Effectively none. This is the reason the premium exists. | Effectively none. |
| What it is | A general cloud that also rents accelerators. Most expensive per hour, and the one your compliance team has already approved. | A general cloud that also rents accelerators. Most expensive per hour, and the one your compliance team has already approved. |
| How you reach it | You get virtual machines. The familiar cloud model. | You get virtual machines. The familiar cloud model. |
| Accelerators | H100 (P5), B200 (P6), L4/L40S (G6), plus Trainium and Inferentia | H100, H200, B200, plus TPU v5e/v6/v7 |
| Regions | Global | Global |
| Pricing model | On-demand, spot, Savings Plans, and Capacity Blocks for reserved GPU windows. | On-demand, spot, committed use discounts, Dynamic Workload Scheduler. |
| Getting started | Existing AWS account. | Existing GCP account. |
| Capacity | Constrained on the newest parts; Capacity Blocks exist precisely because on-demand cannot be relied on. | Constrained on newest parts; DWS is the queueing mechanism. |
| Ownership | Amazon. | Alphabet. |
AWS (P5, P6, G6)
For: Anyone whose data, VPC, compliance boundary and team already live in AWS. The GPU price is rarely the deciding number.
The catch: Reported around $9.36/GPU-hour for a B200 Capacity Block against roughly $5.50 at Lambda. You are buying integration and counterparty certainty, and paying for both.
Economics: Egress and adjacency costs usually dominate the GPU line. Compare total workload cost, never the hourly rate alone.
Google Cloud (A3, A4)
For: Teams already on GCP, and anyone who wants TPUs — which exist nowhere else.
The catch: TPU is the real differentiator and the real lock-in. Choosing a TPU is choosing Google Cloud for the life of that workload; there is no second supplier.
Economics: Spot B200 reported around $6.69/GPU-hour. Committed-use discounts change the picture substantially and are where the actual negotiation happens.
Neither table row is a price
Deliberately. Published on-demand rates at this layer move weekly, and essentially nobody signing a real contract pays them — every serious buyer pays less than every list figure either of these companies publishes. Quoting one here would date this page within a month.
The GPU rental price index carries dated, sourced figures instead, and the durable finding there is the spread: the identical H100 rents from roughly $1.38 to $12.29 an hour depending only on who you rent it from.
The layers underneath both
Whichever you pick is renting you chips in a building that needs power. In 2026 that is the constraint that binds: Microsoft has disclosed an Azure backlog it cannot fill for want of megawatts rather than accelerators, and the US interconnection queue exceeds 2,600 GW with roughly 80% of projects withdrawing before they energise.
Layer 2 — Silicon · Layer 1 — Energy · The interconnection queue