GCP Compute Instance

m4n-hypermem-32

The m4n-hypermem-32 is a memory-optimized GCP Compute instance with 32 vCPUs and 496 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 $5.1329/hr in us-central1. Spot pricing starts at $2.0532/hr in us-central1.

Price History

Pricing by Region

Region On-Demand ↑Spot Reserved 1Y
us-central1$5.133$2.053$3.234
us-east1$5.133$2.053$3.234
us-east4$5.133$2.053$3.234
us-east5$5.133$2.053$3.234
us-west1$5.133$2.053$3.234
asia-south1$5.338$2.135$3.363
asia-south2$5.338$2.135$3.363
europe-west4$5.389$2.156$3.395
europe-west1$5.652$2.261$3.561
northamerica-northeast1$5.652$2.261$3.561
northamerica-northeast2$5.652$2.261$3.561
us-west4$5.781$2.313$3.642
europe-west2$5.852$2.341$3.686
europe-southwest1$6.057$2.423$3.816
europe-west3$6.057$2.423$3.816
europe-central2$6.211$2.484$3.913
asia-southeast1$6.332$2.533$3.989
australia-southeast1$6.416$2.566$4.042
europe-west12$6.622$2.649$4.172
southamerica-west1$7.340$2.936$4.624
southamerica-east1$8.151$3.260$5.135
me-central2$8.213$3.285$5.174

Spot Interruption Risk

Frequently Asked Questions

What is the m4n-hypermem-32?
The m4n-hypermem-32 is a memory-optimized GCP Compute instance with 32 vCPUs and 496 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 $5.1329/hr in us-central1. Spot pricing starts at $2.0532/hr in us-central1.
What is the cheapest region for m4n-hypermem-32?
us-central1 at $5.133/hr on-demand.
How much does m4n-hypermem-32 cost per month?
Starting at $3747.04/mo on-demand in us-central1, or $1498.81/mo with spot pricing.