Instance Plans
Spot
Dynamic compute pricing based on market prices
On Demand
Fixed monthly cost for compute
Compute Flavors
General Purpose
Balanced performance for a wide variety of workloads.
Compute Heavy
High performance for CPU-intensive applications.
Memory Heavy
Large memory allocation for data-intensive applications.
GPU
New
NVIDIA H100 and A30 GPUs for AI and ML workloads.
Pricing
Resources
Documentation
All the details you need to learn about spot.
Blog
The latest industry news, updates and info.
Case Studies
Learn how our customers are making big changes.
Status Page
Real-time system performance for spot infrastructure
Community Showcase
Explore innovative projects built by the Spot community
Changelog
Login
Log in
Get Started
Create a Cloudspace
Instances
›
AWS
›
trn1
Machine Learning ASIC
trn1
2 instance sizes in this family
trn1.2xlarge
8 vCPU · 32 GiB · 1.3438/hr
trn1.32xlarge
128 vCPU · 512 GiB · 21.5/hr
