M3 family · Memory Optimizedx86_64 · Intel/AMD

m3-ultramem-128

The m3-ultramem-128 machine type has 128 vCPUs and 3904 GB of memory. Pricing for this instance starts at $24.36 per hour and $17786.30 monthly in the us-central1 region.

Updated June 21, 2026
vCPUs
128
Memory
3904 GB
Network
Up to 32 Gbps
Storage
Local SSD

Save up to $8893/mo on every m3-ultramem-128

Connect Google Cloud — we'll find the optimal pricing model for your workload.

All pricing options

us-central1 · USD · monthly = 730 hrs
Prices exclude local taxes
TermHourlyMonthlySavings
On-Demand
$24.3648$17786.30 Baseline
1-Year Committed Use Discount (CUD)
$14.4205$10526.95 -41%
3-Year Committed Use Discount (CUD)
$7.3107$5336.83 -70%
Preemptible / Spot
$12.1824$8893.15 -50%
You could save $8893.15/mo on this instance with the right pricing model.
Find my savings

Monthly cost estimator

Estimate your spend based on actual usage.
730 h
1h365h730h (24/7)
1
Estimated monthly
$17786.30
730 hrs × $24.3648/hr × 1

Compute

x86_64 · Intel/AMD
vCPUs i128
Memory i3904 GB
Physical processorIntel Ice Lake
Nested VirtualizationNot supported
Sole TenantNot supported
GPUNone
Within M3 family
m3-ultramem-32
32 vCPU
m3-megamem-64
64 vCPU
m3-ultramem-64
64 vCPU
m3-megamem-128
128 vCPU
m3-ultramem-128
128 vCPU

Lock in this rate across your fleet — typically save 30–40%

Connect your Google Cloud account to apply this pricing logic to every running instance automatically.

Networking

Max egress bandwidth32 Gbps
Tier 1 bandwidth100 Gbps
Enhanced networking iYes (gvnic/virtio)
IPv6 supportYes (Dual-stack VPC)

Storage

Max persistent disks128
Max disk size512 TB
Local SSDsSupported (Scratch Disk)

Compare with another instance

vs.
Specm3-ultramem-128m3-ultramem-32Δ
vCPUs12832-75%
Memory3904 GB976 GB-75%
Hourly Price$24.3648$6.0912-75%
Monthly Price$17786.30$4446.58-75%
Cost optimization

GCP recommender alerts can cut this bill ~30-40%.

Connect your GCP account. We'll identify every underutilized m3.* instance and show exactly where downsizing, cleaning up orphaned disks, or shifting to CUDs saves money — with zero code changes.