Layer 2.5 · head to head
Hyperstack (NexGen Cloud) vs DataCrunch
The middle band: cheaper than hyperscalers, more accountable than marketplaces.
| Hyperstack (NexGen Cloud) | DataCrunch | |
|---|---|---|
| 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 | Small, private, young. | Small European provider. |
| What it is | Purpose-built GPU cloud. Owns or leases its own datacentre capacity and sells it directly. | Purpose-built GPU cloud. Owns or leases its own datacentre capacity and sells it directly. |
| How you reach it | You get virtual machines. The familiar cloud model. | You get virtual machines. The familiar cloud model. |
| Accelerators | H100, H200, A100, L40S | H100, H200, B200, A100, V100 |
| Regions | EU, US, Canada | Finland, Iceland |
| Pricing model | Hourly on-demand and reserved, published. | Hourly and fixed-term contracts, published. |
| Getting started | Self-serve. | Self-serve. |
| Capacity | Moderate and growing. | Small but real; renewable-powered Nordic sites. |
| Ownership | Private. | Private. |
Hyperstack (NexGen Cloud)
For: Cost-sensitive teams who still want a real datacentre rather than a marketplace host.
The catch: Less well known, so less community knowledge when something breaks at 2am.
Economics: Prices meaningfully below the hyperscalers, above the marketplaces. That middle band is a legitimate place to sit.
DataCrunch
For: European teams wanting low-cost datacentre GPUs with a clean power story.
The catch: Two-region footprint. Latency and redundancy are what you trade for the price.
Economics: Among the cheaper genuine-datacentre options in Europe.
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