GCP Compute Instance

m4n-hypermem-64

The m4n-hypermem-64 is a memory-optimized GCP Compute instance with 64 vCPUs and 992 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 $10.2659/hr in us-central1. Spot pricing starts at $4.1063/hr in us-central1.

Price History

Pricing by Region

Region On-Demand ↑Spot Reserved 1Y
us-central1$10.266$4.106$6.468
us-east1$10.266$4.106$6.468
us-east4$10.266$4.106$6.468
us-east5$10.266$4.106$6.468
us-west1$10.266$4.106$6.468
asia-south1$10.676$4.271$6.726
asia-south2$10.676$4.271$6.726
europe-west4$10.779$4.312$6.791
europe-west1$11.304$4.522$7.121
northamerica-northeast1$11.304$4.522$7.121
northamerica-northeast2$11.304$4.522$7.121
us-west4$11.563$4.625$7.284
europe-west2$11.703$4.681$7.373
europe-southwest1$12.114$4.846$7.632
europe-west3$12.114$4.846$7.632
europe-central2$12.421$4.969$7.826
asia-southeast1$12.665$5.066$7.979
australia-southeast1$12.833$5.133$8.084
europe-west12$13.243$5.297$8.343
southamerica-west1$14.680$5.872$9.249
southamerica-east1$16.301$6.520$10.270
me-central2$16.426$6.570$10.348

Spot Interruption Risk

Frequently Asked Questions

What is the m4n-hypermem-64?
The m4n-hypermem-64 is a memory-optimized GCP Compute instance with 64 vCPUs and 992 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 $10.2659/hr in us-central1. Spot pricing starts at $4.1063/hr in us-central1.
What is the cheapest region for m4n-hypermem-64?
us-central1 at $10.266/hr on-demand.
How much does m4n-hypermem-64 cost per month?
Starting at $7494.08/mo on-demand in us-central1, or $2997.63/mo with spot pricing.