Thunder Compute is active here, founded in 2024, headquarters Cambridge, Massachusetts. The tagline is low-cost cloud GPUs with transparent pricing. Billing is per hour. The recorded payment method is credit card. The published rows are RTX A6000, L40, A100 80GB, and H100 SXM.
Why Thunder Compute stands out
Our data records persistent storage, multi-GPU, an API, Jupyter, Docker, and SOC 2. Kubernetes, spot, and reserved instances are not set. “Low-cost” is the tagline. The pricing-competitiveness score is only a little above the middle of this directory, so use the hourly table.
Pros
- Published A6000, L40, A100 80GB, and H100 SXM
- Per-hour billing and credit card
- Jupyter, Docker, an API, multi-GPU, and persistent storage recorded
- Marked SOC 2 compliant
Cons
- No Kubernetes, spot, or reserved instances in our data
- Four published GPUs
- Founded in 2024
Getting started
- Open Thunder Compute and compare the H100 or A100 hourly row.
- Pay by credit card.
- Use Jupyter or Docker. Kubernetes and spot are not recorded.
Best for: An hourly H100 SXM or A100 with Jupyter and a SOC 2 flag, from a young Cambridge provider, without Kubernetes or spot.