Announced at Google Cloud Next on 22 April 2026 and not yet generally available. Every figure comes from Google's announcement blog rather than from a spec table for a running machine type. Captive silicon: Google has never sold TPUs as hardware and no third party can resell or host them, so this pod cannot be bought and Google Cloud will be the only place to rent one.

Google TPU 8t Superpod

pod
9600× TPU 8t
3D torus ICI with 47 Pb/s non-blocking bisectional bandwidth; Virgo network links over 134,000 chips
2026
FP4
121
EFLOPS
Power
—
kW Total
Memory
2073600
GB Total

Google's eighth-generation training superpod: 9,600 TPU 8t chips on a 3D torus, with 47 petabits per second of non-blocking bisectional bandwidth, and a Virgo network able to link over 134,000 chips across superpods. At Google's published 12.6 PFLOPS of FP4 per chip that is 120.96 EFLOPS across the pod, with 2.07 PB of HBM3e and 62.7 PB/s of aggregate memory bandwidth. Google publishes no pod-level compute figure for this generation; the total here is chip count times its published per-chip peak, which is Google's own arithmetic — on the two generations where it publishes both, v4 and Ironwood, the pod figure equals exactly that product. FP4 is the only precision Google states, and it publishes no sparse figure for any TPU, so this is dense. Announced at Google Cloud Next in April 2026 and not yet generally available.

We will point you at suppliers who have it. Free, and no signup.

FP4
121
EFLOPS

System Details

GPU Configuration

GPU Model: Google TPU 8t
GPU Count: 9600 GPUs
Architecture: TPU 8
Interconnect: 3D torus ICI with 47 Pb/s non-blocking bisectional bandwidth; Virgo network links over 134,000 chips

System Specifications

Form Factor: pod
Total Power: —
Total Memory: 2.1 PB
On-die SRAM: 1.23 TB
Memory Bandwidth: 62.7 PB/s

Precision Performance Breakdown

PrecisionSystem PerformancePer GPUEfficiency
121 EFLOPS 12,600 TFLOPS —

Where a vendor states a basis, the figure shown is the dense one, and any figure it publishes is listed separately. Vendors commonly headline the sparse number instead: for NVIDIA's 2:4 structured sparsity that is exactly twice the dense figure, though other vendors' sparse modes do not all follow that ratio.

Powered by Google TPU 8t

This system utilizes 9600 × Google TPU 8t GPUs, each delivering exceptional performance for AI and HPC workloads.

Per GPU TDP

—

Per GPU Memory

216 GB

Process Node

—

Architecture

TPU 8

Documentation & Resources

Flopper Spec Sheet

Google TPU 8t Superpod specifications, generated from our database. Printable.

Vendor product page

Google TPU 8t Superpod documentation

View ↗

Cloud TPU v5p

Google • 2023-12-07

View Document ↗

Typical Use Cases

AI/ML Training
High-Performance Computing
Data Analytics

The Google TPU 8t Superpod runs 9600× TPU 8t GPUs over 3D torus ICI with 47 Pb/s non-blocking bisectional bandwidth; Virgo network links over 134,000 chips, delivering 121 EFLOPS FP4 dense.

© 2026 Flopper.io - Compare the hardware powering AI