About Hyperdisk Balanced
This document describes the features of Hyperdisk Balanced, the general-purposeHyperdisk type. Hyperdisk Balanced is the best fit for most workloads andoffers the best combination of price and performance for total cost ofownership. For more information about Hyperdisk and the otherHyperdisk types, seeAbout Hyperdisk.
You can specify up to 160,000 IOPSand 2,400 MiB/s of throughput for a single Hyperdisk Balanced volume.Hyperdisk Balanced is designed for sub-millisecond latencies.
To create a new Hyperdisk Balanced volume, seeCreate a Hyperdisk volume.
Use cases
Hyperdisk Balanced is a good fit for most workloads, including the followinguse cases:
- Databases, like MySQL and PostgresQL
- Virtual Desktop Infrastructure (VDI)
- Line of Business (LOB) applications
- Boot disks for compute instances: Hyperdisk Balanced and Hyperdisk Balanced High Availability arethe only supported Hyperdisk types for all instances, includingfor TPU VMs
Machine series support
You can use Hyperdisk Balanced with the following machine series:
About provisioned performance for Hyperdisk Balanced
You don't have to provision performance when you createHyperdisk volumes. If you don't provision performance, Compute Enginecreates the volume with default values that you can modify later.For details about default values, seeDefault IOPS andthroughput values.
If you know your performance needs, you can specify IOPS and throughput limitsfor a Hyperdisk Balanced volume when you create the volume, and you can changethe provisioned values after you create the volume. You can't specify anIOPS or throughput level if you don't specify a size.
Important: Hyperdisk volumescan't reach the provisioned performance unless the compute instance supports thatlevel of performance. For detailed performance limits for all supported instancesby machine type, seePerformance limits when attached toan instance.Size and performance limits
The following limits apply to the size, IOPS, and throughput values you canspecify for a Hyperdisk Balanced volume.
- Size: between 4 GiB and64 TiB. The default size is100 GiB.
- IOPS: between 3,000 and 160,000 IOPS.For volumes less than 320 GiB, the range of IOPS values you can specifyvaries by size. For details, seeLimits for provisioned IOPS.
- Throughput: between 140 and 2,400 MiB/s of throughput, butthe throughput value depends on how much IOPS you provision. For details,seeLimits for provisioned throughput.
Limits for provisioned IOPS
You can choose the IOPS limit for a Hyperdisk Balanced volume according to thefollowing limits. The minimum and maximum limits vary by size.
- Minimum configurable IOPS:
- 4 or 5 GiB volumes: IOPS is fixed at 2,000 and2,500 IOPS, respectively.
- 6 GiB-64 TiB: at least 3,000IOPS.
- Maximum configurable IOPS:
- 6-319 GiB volumes: the maximum IOPS limit depends on the volume'ssize. See the following table for more information.
- 320 GiB-64 TiB volumes: at most160,000 IOPS.
To calculate the range of configurable IOPS values for a specific volume size,use these formulas:
- Minimum IOPS: 3,000 IOPS
- Maximum IOPS:
MIN (500x, 160,000)MiB/s, wherexis the volume's size.
The following table lists the limits for configurable IOPS for common volumesizes.
| Size (GiB) | Configurable IOPS values |
|---|---|
| 4 | 2,000 |
| 5 | 2,500 |
| 10 | 3,000 to 5,000 |
| 50 | 3,000 to 25,000 |
| 150 | 3,000 to 75,000 |
| 319 | 3,000 to 159,500 |
| 320 or more | 3,000 to 160,000 |
Limits for provisioned throughput
Volumes with 4 or 5 GiB of capacity have a fixed throughput value of140 MiB/s. For larger volumes, you can provision additionalthroughput. The range of supported throughput values depends on the provisionedIOPS and is calculated with the following formulas, whereP is the provisionedIOPS.
- Minimum throughput:
MAX (140, P/256)MiB/s - Maximum throughput:
MIN (2,400, P/4)MiB/s
The following table lists the limits for some provisioned IOPS values and thecorresponding throughput ranges. If a size isn't listed, use the precedingformula to calculate the allowable values.
| Provisioned IOPS | Configurable throughput (MiB/s) |
|---|---|
| 2,000 | 140 to 500 |
| 2,500 | 140 to 625 |
| 3,000 | 140 to 750 |
| 4,000 | 140 to 1,000 |
| 8,000 | 140 to 2,000 |
| 32,000 | 140 to 2,400 |
| 50,000 | 156 to 2,400 |
| 64,000 | 250 to 2,400 |
| 150,000 | 585 to 2,400 |
| 160,000 | 625 to 2,400 |
Default size, IOPS, and throughput values
If you don't specify a size, IOPS, or throughput value when you create a Hyperdisk Balancedvolume, Compute Engine assigns default values.
The default size for Hyperdisk Balanced volumes is 100 GiB.
The default IOPS and throughput are based on the following formulas, wherex is the volume's size in GiB.
- Default IOPS:
- Volumes 6 GiB or less:
500xIOPS - Volumes between 6 GiB and 26.666667 TiB:
6x + 3,000IOPS - Volumes larger than 26.666667 TiB: 160,000 IOPS
- Volumes 6 GiB or less:
- Default throughput:
- Volumes 6 GiB or less:
140MiB/s - Volumes larger than 6 GiB:
1.5x + 140 MiB/s
- Volumes 6 GiB or less:
Baseline performance
Hyperdisk Balanced volumes include baseline, or free performance. The first 3,000 IOPS and140 MiB/s of throughput that you provision for a volume are free.Performance you provisionabove the baseline amounts isbillable performance,and is not free. If you provision a volume with 5,000 IOPS, you're billed for2,000 IOPS. Similar principles apply to throughput.
Baseline performance is also included with Hyperdisk Balanced volumes that you create in thefollowing Hyperdisk Storage Pools:
- Standard performance Hyperdisk Storage Pools
- Standard capacity Hyperdisk Storage Pools
- Advanced capacity Hyperdisk Storage Pools
Hyperdisk Balanced volumes in Advanced performance Hyperdisk Storage Pools don't include baselineperformance.
Baseline performance and quota
Baseline IOPS and throughput don't affect the IOPS and throughput quota for aproject's Hyperdisk Balanced and Hyperdisk Balanced High Availability volumes.
For example, if your project has a quota of 500,000 IOPS and you create a Hyperdisk Balancedvolume with 3,000 IOPS, then you still have a quota of 500,000 IOPSafter you create the volume. If you also create a Hyperdisk Balanced High Availability volume with 4,000 IOPS,then the project would have a remaining IOPS quota of 499,000.
Limits for consuming baseline performance
A project's Hyperdisk Balanced and Hyperdisk Balanced High Availability volumesthat are in the same zone can concurrently use up to 50 GiB/sof baseline throughput and up to 500,000 of baseline IOPS.This limit is referred to as theconcurrent consumption limit for baseline performance.
The concurrent consumption limit for baseline performance restricts how much baseline performance thevolumes canuse at the same time, and doesn't affect much baseline performanceyou can provision. The concurrent consumption limit doesn't apply to consuming orprovisioning billable performance.
For more information, seeConcurrent consumption limits for baseline performance.
Change the provisioned performance or size
You can change the provisioned size of a Hyperdisk Balanced volume atmost twice within a 4-hour window, and you can change the provisionedperformance at most once every 4 hours. To learn howto modify size or performance, seeModify a Hyperdisk volume.
Performance limits when attached to an instance
This section lists the performance limits for each machine type that supportsHyperdisk Balanced. You can specify up to 160,000 IOPSand 2,400 MiB/s of throughput for a single Hyperdisk Balanced volume.
This section lists the maximum performance that Hyperdisk Balanced volumescan achieve for each supported instance. A Hyperdisk Balanced volume's performancewhen it's attached to an instance can't exceed the limits for the instance'smachine type.The performance limits are also shared across all Hyperdisk Balanced volumesattached to the same instance, regardless of each volume's provisioned performance.
Achieve higher performance with multiple Hyperdisk Balanced volumes
Certain instances can exceed the maximum performance for a single Hyperdisk Balanced volume—160,000 IOPS or 2,400 MiB per second (MiB/s). An instance can onlyachieve these higher limits if you attach multiple Hyperdisk Balanced volumes to the instance.
If an instance's performance limit in the following table exceeds 160,000 IOPS or2,400 MiB/s, then the instance can only achieve that limitif you attach multiple Hyperdisk Balanced volumes.For example, the performance limit for ac4a-*-72 instance using Hyperdisk Balancedis 240,000 IOPS and 5,000 MiB/s. To achieve this maximum performanceyou must attach at least two Hyperdisk Balanced volumes to the instance.
Hyperdisk performance is half-duplex,so the IOPS and throughput limits are shared between read and write operations.This means that if a machine type has a 50,000 IOPS limit, then the sum of the readsand writes every second can't exceed 50,000.
The performance limits also apply tocustom machine types.
| Instance machine type | Maximum IOPS | Maximum throughput (MiB/s) |
|---|---|---|
| A3 (A3+H100) | ||
a3-megagpu-8g | 160,000 | 4,800 |
a3-highgpu-1g | 25,000 | 400 |
a3-highgpu-2g | 120,000 | 1,800 |
a3-highgpu-4g | 160,000 | 2,400 |
a3-highgpu-8g | 160,000 | 4,800 |
a3-edgegpu-8g | 160,000 | 4,800 |
| A3 (A3+H200) | ||
a3-ultragpu-8g2 | 320,000 | 10,000 |
| A42 | ||
a4-highgpu-8g2 | 320,000 | 10,000 |
| A4X2 | ||
a4x-highgpu-4g2 | 320,000 | 10,000 |
| C3 | ||
c3-*-4 | 25,000 | 400 |
c3-*-8 | 50,000 | 800 |
c3-*-22 | 120,000 | 1,800 |
c3-*-44 | 160,000 | 2,400 |
c3-*-88 | 160,000 | 4,800 |
c3-*-176 | 160,000 | 10,000 |
c3-*-192 | 160,000 | 10,000 |
| C3D | ||
c3d-*-4 | 25,000 | 400 |
c3d-*-8 | 50,000 | 800 |
c3d-*-16 | 75,000 | 1,200 |
c3d-*-30 | 75,000 | 1,200 |
c3d-*-60 | 160,000 | 2,400 |
c3d-*-90 | 160,000 | 2,400 |
c3d-*-180 | 160,000 | 2,400 |
c3d-*-360 | 160,000 | 2,400 |
| C42 | ||
c4-*-22 | 50,000 | 400 |
c4-*-42 | 50,000 | 400 |
c4-*-82 | 50,000 | 800 |
c4-*-162 | 100,000 | 1,600 |
c4-*-242 | 100,000 | 1,600 |
c4-*-322 | 100,000 | 1,600 |
c4-*-482 | 160,000 | 2,400 |
c4-*-962 | 240,000 | 4,800 |
c4-*-1442 | 240,000 | 4,800 |
c4-*-1922 | 320,000 | 10,000 |
c4-*-2881,2 | 320,000 | 10,000 |
| C4A | ||
c4a-*-1 | 25,000 | 400 |
c4a-*-2 | 50,000 | 800 |
c4a-*-4 | 50,000 | 800 |
c4a-*-8 | 50,000 | 1,000 |
c4a-*-16 | 60,000 | 1,600 |
c4a-*-32 | 120,000 | 2,400 |
c4a-*-48 | 160,000 | 3,300 |
c4a-*-64 | 240,000 | 4,400 |
c4a-*-72 | 240,000 | 5,000 |
| C4D2 | ||
c4d-*-22 | 22,500 | 400 |
c4d-*-42 | 50,000 | 400 |
c4d-*-82 | 50,000 | 800 |
c4d-*-162 | 75,000 | 1,200 |
c4d-*-322 | 75,000 | 1,600 |
c4d-*-482 | 75,000 | 1,600 |
c4d-*-642 | 160,000 | 2,400 |
c4d-*-962 | 160,000 | 2,800 |
c4d-*-1922 | 240,000 | 4,800 |
c4d-*-3842 | 320,000 | 10,000 |
| G4 | ||
g4-standard-48 | 75,000 | 1,600 |
g4-standard-96 | 160,000 | 2,800 |
g4-standard-192 | 240,000 | 4,800 |
g4-standard-384 | 320,000 | 10,000 |
| H3 | ||
h3-standard-88 | 15,000 | 240 |
| H4D2 | ||
h4d-*-192 | 15,000 | 240 |
| M1 | ||
m1-ultramem-40 | 60,000 | 1,200 |
m1-ultramem-80 | 100,000 | 2,400 |
m1-megamem-96 | 100,000 | 4,000 |
m1-ultramem-160 | 100,000 | 4,000 |
| M2 | ||
m2-*-208 | 100,000 | 2,400 |
m2-*-416 | 100,000 | 4,000 |
| M3 | ||
m3-*-32 | 160,000 | 2,400 |
m3-*-64 | 160,000 | 4,800 |
m3-*-128 | 160,000 | 4,800 |
| M42 | ||
m4-*-162 | 100,000 | 1,600 |
m4-*-282 | 100,000 | 1,600 |
m4-*-322 | 120,000 | 1,800 |
m4-*-562 | 160,000 | 2,400 |
m4-*-642 | 180,000 | 2,800 |
m4-*-1122 | 240,000 | 4,800 |
m4-*-2242 | 320,000 | 10,000 |
| N4 | ||
n4-*-2 | 15,000 | 240 |
n4-*-4 | 15,000 | 240 |
| Custom N4 types with 6 vCPUs | 15,000 | 240 |
n4-*-8 | 15,000 | 480 |
| Custom N4 types with 10 to 14 vCPUs | 15,000 | 480 |
n4-*-16 | 80,000 | 1,200 |
| Custom N4 types with 18 to 30 vCPUs | 80,000 | 1,200 |
n4-*-32 | 100,000 | 1,600 |
| Custom N4 types with 34 to 46 vCPUs | 100,000 | 1,600 |
n4-*-48 | 160,000 | 2,400 |
| Custom N4 types with 50 to 62 vCPUs | 160,000 | 2,400 |
n4-*-64 | 160,000 | 2,400 |
| Custom N4 types with 66 to 78 vCPUs | 160,000 | 2,400 |
n4-*-80 | 160,000 | 2,400 |
| N4A(Preview) | ||
n4a-*-1 | 15,000 | 240 |
n4a-*-2 | 15,000 | 240 |
| Custom N4A types with 3 vCPUs | 15,000 | 240 |
n4a-*-4 | 15,000 | 240 |
| Custom N4A types with 5 to 7 vCPUs | 15,000 | 240 |
n4a-*-8 | 15,000 | 480 |
| Custom N4A types with 9 to 15 vCPUs | 15,000 | 480 |
n4a-*-16 | 80,000 | 1,200 |
| Custom N4A types with 17 to 31 vCPUs | 80,000 | 1,200 |
n4a-*-32 | 100,000 | 1,600 |
| Custom N4A types with 33 to 47 vCPUs | 100,000 | 1,600 |
n4a-*-48 | 160,000 | 2,400 |
| Custom N4A types with 49 to 63 vCPUs | 160,000 | 2,400 |
n4a-*-64 | 160,000 | 2,400 |
| N4D | ||
n4d-*-2 | 30,000 | 240 |
n4d-*-4 | 30,000 | 240 |
n4d-*-8 | 30,000 | 480 |
n4d-*-16 | 80,000 | 1,200 |
n4d-*-32 | 100,000 | 1,600 |
n4d-*-48 | 160,000 | 2,400 |
n4d-*-64 | 160,000 | 2,400 |
n4d-*-80 | 160,000 | 2,400 |
n4d-*-96 | 160,000 | 2,400 |
| TPU v6e | ||
ct6e-standard-1t | 75,000 | 1,200 |
ct6e-standard-4t | 160,000 | 2,400 |
ct6e-standard-8t | 160,000 | 2,400 |
| X4 | ||
x4-480-*-metal | 160,000 | 4,800 |
x4-960-*-metal | 160,000 | 4,800 |
x4-1440-24t-metal | 160,000 | 4,800 |
x4-1920-32t-metal | 160,000 | 4,800 |
| Z3 | ||
z3-*-8 | 50,000 | 800 |
z3-*-14 | 100,000 | 1,600 |
z3-*-16 | 100,000 | 1,600 |
z3-*-22 | 120,000 | 1,800 |
z3-*-32 | 160,000 | 2,400 |
z3-*-44 | 160,000 | 2,400 |
z3-*-88 | 160,000 | 4,800 |
z3-*-176 | 160,000 | 4,800 |
z3-*-1921 | 160,000 | 4,800 |
1Includes bare metal instances.
2Also offers steady state performance. For more information, seePerformance limits for machine types with steady state performance.
Disaster recovery protection for Hyperdisk Balanced volumes
You can back up a Hyperdisk Balanced volume withinstant snapshotsandstandard snapshots.Snapshots back up the data on a Hyperdisk Balanced volume at a specific point in time.
Replication across multiple regions
You can protect your data in the unlikely event of a regional outage by enabling Asynchronous Replication. Asynchronous Replication maintains a copy ofthe data on your volume in another region. For example, to protect aHyperdisk Balanced volume inus-west1,you can use Asynchronous Replication to replicate the volume to a secondary volume intheus-east4 region. If the volume inus-west1became unavailable, then you could use the secondary volume inus-east4.
Cross-zonal replication
You can't replicate Hyperdisk Balancedvolumes to another zone. To replicate data to another zone within the same region,you must useHyperdisk Balanced High Availability volumes.
Regional availability for Hyperdisk Balanced
Hyperdisk Balanced is available in all zones and regions.
Hyperdisk Balanced Storage Pools
To simplify planning, avoid overprovisioning storage, and reduce costs, considerusing Hyperdisk Balanced Storage Pools. Hyperdisk Balanced Storage Pools let you purchase Hyperdisk Balanced storage andperformance in bulk. To learn more, seeAbout Hyperdisk Storage Pools.
Share a Hyperdisk Balanced volume between VMs
You can use multi-writer mode to attach a Hyperdisk Balanced volume to multipleinstances, while each instance maintains write access. This feature enables highavailability for workloads like Microsoft SQL Server Failover ClusterInfrastructure.
To learn more, seeShare a disk between VMs.
Pricing
You are billed for the total provisioned size, IOPS, and throughput of yourHyperdisk Balanced volumes until you delete them. Charges incur even if thevolume isn't attached to any instances or if the instance is suspended orstopped. For more information seeDisk pricing.
Limitations for Hyperdisk Balanced
- Hyperdisk Balanced volumes are zonal and can only be accessed from thezone where the volume was created. If you need to access a Hyperdiskvolume from more than one zone, create a Hyperdisk Balanced High Availability volume.
- You can't create amachine imagefrom a Hyperdisk volume.
- Hyperdisk Balanced volumes in multi-writer mode have additionallimitations.
- Hyperdisk Balanced volumes in Confidential mode have additionallimitations.
- You can change a Hyperdisk Balanced volume's size and performance every4 hours.
- Compute Engine enforces a concurrent consumption limit for baseline performance for a project's Hyperdisk Balancedand Hyperdisk Balanced High Availability volumes that are in the same zone. The limits are 500,000 IOPS and 50 GiB/s ofthroughput. For more information, seeConcurrent consumption limits for baseline performance.
What's next
Except as otherwise noted, the content of this page is licensed under theCreative Commons Attribution 4.0 License, and code samples are licensed under theApache 2.0 License. For details, see theGoogle Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2025-12-15 UTC.