TPU v5e — specs & cloud pricing
The TPU v5e is a datacenter accelerator from Google. Live specs and modeled cloud $/hr from the Mintok silicon catalog.
Cloud rate
$1.20per chip-hour (modeled)
Specifications
- Vendor
- Class
- Datacenter accelerator
- Architecture
- TPU
- BF16 throughput
- 197 TFLOPS
- On-package memory
- 16 GB
- Memory bandwidth
- 819 GB/s
- TDP
- 250 W
More Google silicon
FAQ
How much does the TPU v5e cost per hour in the cloud?
The TPU v5e rents for about $1.20 per chip-hour on cloud providers (modeled from public rates). Own-vs-rent economics — capex, lease and TCO — are in the Mintok sizing report and rack teardowns.
How much memory does the TPU v5e have?
The TPU v5e has 16 GB of on-package memory at 819 GB/s bandwidth.
What is the TPU v5e's BF16 compute throughput?
The TPU v5e delivers 197 BF16 TFLOPS at a 250 W TDP.
Is the TPU v5e a datacenter or edge accelerator?
The TPU v5e is a datacenter accelerator (250 W TDP).
Cost this chip in a fleet
Size a deployment on the TPU v5e, or see the full priced rack BOM it sits in.
Specs and modeled rates maintained by Mintok's catalog watchers from public sources. Not investment advice. Methodology.