Source: https://www.tera.gw/guides
Retrieved: 2026-11-21T21:36:04.769Z (page retrieval, not inventory verification)
Content SHA-256: 261a8b228fc0d68e74fc43801cf83d633f3266c8c2363c56fcf3874cbd756e87

# Guides

Free guides and tools for GPU rental and colocation.

## GPU rental

- ### [GPU rental pricing](<https://www.tera.gw/pricing>)

Indicative offers for dedicated GPU nodes: contract term, rate per GPU-hour, prepayment and location.

- ### [GPU rental cost calculator](<https://www.tera.gw/guides/gpu-rental-cost-calculator>)

The cost of a cluster for each hour, each month and the full term, and the prepayment.

- ### [Renting vs. buying GPUs for an AI startup](<https://www.tera.gw/guides/renting-vs-buying-gpus>)

Compare rental commitments with hardware ownership, hosting and operating costs before choosing how to fund your next GPU deployment.

- ### [How dedicated GPU rental works](<https://www.tera.gw/guides/how-dedicated-gpu-rental-works>)

From request to running nodes: what a dedicated GPU rental includes, how it is priced and what to confirm in writing.

- ### [Bare metal vs. cloud GPU instances](<https://www.tera.gw/guides/bare-metal-vs-cloud-gpus>)

Dedicated bare-metal nodes and on-demand cloud instances suit different workloads. Compare control, commitment and cost.

- ### [Contract term and prepayment in a GPU rental](<https://www.tera.gw/guides/gpu-rental-term-and-prepayment>)

How the term, the GPU-hour rate and the prepayment work together, and how to compare two offers on the same basis.

## GPU hardware

- ### [B300, B200, H200 and H100 compared](<https://www.tera.gw/guides/b300-b200-h200-h100-compared>)

The four NVIDIA data-center GPUs that Tera rents, by generation and by memory for each GPU and for each 8-GPU node.

- ### [InfiniBand vs. RoCE for a GPU cluster](<https://www.tera.gw/guides/infiniband-vs-roce>)

Both connect GPU nodes with low delay for training across nodes. How they differ, and when separate nodes are enough.

- ### [How to size a GPU cluster](<https://www.tera.gw/guides/how-to-size-a-gpu-cluster>)

A method to estimate the number of GPUs and nodes for training, fine-tuning or inference before you ask for an offer.

## Colocation

- ### [GPU colocation](<https://www.tera.gw/colocation>)

Rack space and power for the GPUs that you own.

- ### [What goes into GPU colocation costs?](<https://www.tera.gw/guides/gpu-colocation-costs>)

Power, cooling, rack density, connectivity and service scope determine GPU colocation pricing. Use this checklist to compare written offers.

- ### [Power and cooling for GPU racks](<https://www.tera.gw/guides/gpu-rack-power-and-cooling>)

What a data center must provide for GPU servers: rack power, redundancy and air or liquid cooling, and the figures to bring.

## Reference

- ### [GPU rental and colocation glossary](<https://www.tera.gw/guides/glossary>)

The terms in a GPU rental offer and in a colocation offer, each in one or two sentences.

- ### [News](<https://www.tera.gw/news>)

Announcements and releases from Tera Compute.

---
Interactive controls above are described, not submitted. Reading does not reserve capacity or authorize payment.
For a GPU rental or a colocation request: https://www.tera.gw/
