Skip to main content
HardwareAI Infrastructurevv4

Google TPU v4

by Google · paid · Last verified 2026-03-17

Google's fourth-generation TPU, used internally to train PaLM, LaMDA, and early Gemini models. Features 32GB HBM2 per chip and an optical circuit-switched ICI for flexible pod topology, enabling massive-scale distributed training.

https://cloud.google.com/tpu/docs/v4
C+
C+Average
Adoption: C+Quality: AFreshness: B+Citations: B+Engagement: F

Specifications

License
Proprietary
Pricing
paid
Capabilities
ai-training, inference, bfloat16-compute, optical-ici
Integrations
jax, tensorflow, pytorch-xla, gcp
Use Cases
large-model-training, research, pretraining
API Available
Yes
Tags
tpu, data-center, training, google, cloud
Added
2026-03-17
Completeness
100%

Index Score

58.5
Adoption
55
Quality
85
Freshness
70
Citations
78
Engagement
0

Put AI to work for your business

Deploy this hardware alongside autonomous AaaS agents that handle tasks end-to-end — no babysitting required.

Explore the full AI ecosystem on Agents as a Service