GCP Compute Instance

m4-hypermem-16

The m4-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 $1.4258/hr in us-central1. Spot pricing starts at $0.1282/hr in us-west4.

Price History

Pricing by Region

Region On-Demand ↑Spot Reserved 1Y
us-central1$1.426$0.855$0.841
us-east1$1.426$0.855$0.841
us-east4$1.426$0.390$0.841
us-east5$1.426$0.145$0.841
us-west1$1.426$0.855$0.841
us-west4$1.426$0.128$0.841
asia-south1$1.469$0.392$0.866
asia-south2$1.469$0.157$0.866
northamerica-northeast1$1.569$0.336$0.925
northamerica-northeast2$1.569$0.162$0.925
europe-west3$1.625$0.567$0.959
europe-west4$1.625$0.662$0.959
europe-west1$1.625$0.807$0.959
europe-southwest1$1.711$0.154$1.009
europe-west2$1.725$0.275$1.018
europe-west12$1.782$0.160$1.052
asia-southeast1$1.782$0.885$1.052
australia-southeast1$1.854$0.486$1.094
europe-central2$2.040$0.184$1.203
southamerica-west1$2.040$0.218$1.203
southamerica-east1$2.267$0.272$1.338
me-central2$2.281$0.235$1.346

Spot Interruption Risk

Frequently Asked Questions

What is the m4-hypermem-16?
The m4-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 $1.4258/hr in us-central1. Spot pricing starts at $0.1282/hr in us-west4.
What is the cheapest region for m4-hypermem-16?
us-central1 at $1.426/hr on-demand.
How much does m4-hypermem-16 cost per month?
Starting at $1040.84/mo on-demand in us-central1, or $93.56/mo with spot pricing.