tpu7x-standard-4t by Google Cloud Platform

tpu7x-standard-4t is a 224 vCPUs, 992 GB RAM, 4 Google TPUs server offered by Google Cloud Platform with 224 vCPUs, 960 GiB of memory and 0 GB of storage. The pricing starts at 17.328 USD per hour.
224 vCPU
960 GiB Memory
4 GPU
Spare SCore
-
(All-cores)
-
(Single-core)

Specifications

Server Metadata

Vendor ID
gcp
Server ID
1741023
Name
tpu7x-standard-4t
Description
224 vCPUs, 992 GB RAM, 4 Google TPUs
Family
tpu7x
Hw Virt
-
Status
active
Observed At
2026-09-16T17:56:47.491955

Availability

REGION / IDSPOTONDEMAND
Council Bluffs (US) / us-central1
17.328 USD/h 48 USD/h
Moncks Corner (US) / us-east1
17.328 USD/h 48 USD/h
Columbus (US) / us-east5
19.2 USD/h 48 USD/h
London (GB) / europe-west2
21.12 USD/h 52.8 USD/h

Processor

vCPUs
224
CPU Allocation
Dedicated
CPU Architecture
x86_64

System Resources and Accelerators

MEMORY
Memory Amount
960 GiB
GPU
GPU Count
4
GPU Memory Min
192 GiB
GPU Memory Total
768 GiB
GPU Manufacturer
Google
GPU Family
TPU
GPU Model
v7x
GPUs
  • Google TPU v7x (Memory amount: 196608)
  • Google TPU v7x (Memory amount: 196608)
  • Google TPU v7x (Memory amount: 196608)
  • Google TPU v7x (Memory amount: 196608)
STORAGE
Storage Size
0 GB
Storages
    NETWORK
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    0 GB/month
    IPv4
    0

    Server Description

    An accelerator-focused server featuring four specialized TPUs, 224 dedicated vCPUs, and 960 GB of RAM for machine learning workloads.

    GPU AcceleratedCompute Optimized

    Google Cloud Platform tpu7x-standard-4t is an x86_64 architecture server designed for specialized acceleration workloads. It features 224 dedicated vCPUs and 960.0 GB of system memory, yielding a ratio of 4.29 GB of RAM per core. The primary hardware highlight is the inclusion of four Google TPU v7x accelerators with a total of 768 GB of VRAM. This server does not include local storage, meaning all data persistence must be handled via network-attached storage. It also includes zero complimentary public IPv4 addresses. Due to the high density of dedicated compute resources and specialized TPU hardware, this server is optimized for machine learning training, deep learning inference, and large-scale data science workloads. The high resource density of this hardware class offers a specialized, cost-efficient platform for TPU-native workloads.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Alternatives

    Servers of the Same Family

    INSTANCEvCPUsMEMORYGPUs
    tpu7x-standard-1t 56240 GiB1

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYGPUs

    tpu7x-standard-4t FAQs