Cloud, GPU & Compute Infrastructure
Compute infrastructure companies rent out the GPU capacity that both frontier labs and everyday builders train and serve models on — the hyperscalers (AWS, Azure, GCP) plus a wave of GPU-cloud specialists (CoreWeave, Lambda, Crusoe, Nebius, RunPod) that emerged specifically to meet AI-era demand the hyperscalers couldn’t fill fast enough. The category matters to builders directly: which provider you pick determines your GPU availability, your per-hour cost, and increasingly your access to the newest chip generation before it’s widely available elsewhere. GROUNDING tracks capacity announcements, pricing shifts, and new regions/chip generations coming online, since a compute shortage or price war here changes what’s economically viable to build one layer up the stack.
At a glance
- 7 tracked companies (4 hand-curated)
- Most recently updated: Yandex Cloud (2026-06-21)
Most actively covered
FAQ
What is the Cloud, GPU & Compute Infrastructure category?
Rents cloud or GPU capacity for training and running AI workloads.
How many companies are in this category?
GROUNDING currently tracks 7 companies in Cloud, GPU & Compute Infrastructure, 4 of which have a hand-curated profile.
How current is this hub?
The most recently updated entry is Yandex Cloud, last mentioned 2026-06-21; the radar refreshes hourly.