GCP Compute Instance

m4n-hypermem-16

The m4n-hypermem-16 is a memory-optimized GCP Compute instance with 16 vCPUs and 248 GB of memory, powered by Intel. Designed for memory-intensive applications like in-memory caches, real-time analytics, and large databases. On-demand pricing starts at $2.5665/hr in us-central1. Spot pricing starts at $1.0266/hr in us-central1.

Price History

Pricing by Region

Region On-Demand ↑Spot Reserved 1Y
us-central1$2.566$1.027$1.617
us-east1$2.566$1.027$1.617
us-east4$2.566$1.027$1.617
us-east5$2.566$1.027$1.617
us-west1$2.566$1.027$1.617
asia-south1$2.669$1.068$1.682
asia-south2$2.669$1.068$1.682
europe-west4$2.695$1.078$1.698
europe-west1$2.826$1.130$1.780
northamerica-northeast1$2.826$1.130$1.780
northamerica-northeast2$2.826$1.130$1.780
us-west4$2.891$1.156$1.821
europe-west2$2.926$1.170$1.843
europe-southwest1$3.029$1.211$1.908
europe-west3$3.029$1.211$1.908
europe-central2$3.105$1.242$1.956
asia-southeast1$3.166$1.267$1.995
australia-southeast1$3.208$1.283$2.021
europe-west12$3.311$1.324$2.086
southamerica-west1$3.670$1.468$2.312
southamerica-east1$4.075$1.630$2.567
me-central2$4.106$1.643$2.587

Spot Interruption Risk

Frequently Asked Questions

What is the m4n-hypermem-16?
The m4n-hypermem-16 is a memory-optimized GCP Compute instance with 16 vCPUs and 248 GB of memory, powered by Intel. Designed for memory-intensive applications like in-memory caches, real-time analytics, and large databases. On-demand pricing starts at $2.5665/hr in us-central1. Spot pricing starts at $1.0266/hr in us-central1.
What is the cheapest region for m4n-hypermem-16?
us-central1 at $2.566/hr on-demand.
How much does m4n-hypermem-16 cost per month?
Starting at $1873.52/mo on-demand in us-central1, or $749.41/mo with spot pricing.