General-purpose machine family for Compute Engine (original) (raw)

This document describes the features of the Compute Engine general-purpose machine family, which has the best price-performance with the most flexible vCPU to memory ratios, and provides features that target most standard and cloud-native workloads.

The general-purpose machine family has predefined andcustom machine types to align with your workload, depending on your requirements.

C4D (Preview) is powered by the fifth generation AMD EPYC Turin processor andTitanium. These machine types have up to 384 vCPUs and 3,024 GB of DDR5 memory, a max-boost frequency of 4.1 GHz, and up to 200 Gbps per VM Tier_1 networking performance. C4D also offers bare metal instances and Local SSD machine types with up to 12 TiB of Titanium SSD.

C4A is powered by Google's first generation Axion custom Arm-based processor built on Arm Neoverse V2. C4A provides machine types with up to 72 vCPUs, 576 GB DDR5 memory, 6 TiB of local Titanium SSD, and up to 100 Gbps with per VM Tier_1 networking performance.

C4 and N4 are powered by the fifth generation Intel Xeon Scalable processor (code-named Emerald Rapids). C4 offers a sustained, all-core turbo frequency of 3.1 GHz, 1.5 TB of DDR5 memory, and supports up to 200 Gbps of per VM Tier_1 networking performance. N4 offers a sustained, all-core turbo frequency of 2.9 GHz, 640 GB of DDR5 memory, and up to 50 Gbps of standard network bandwidth.

C3 is powered by fourth generation Intel Xeon Scalable processors and offers a sustained, all-core turbo frequency of 3.0 GHz, 8 channels of DDR5 memory, and up to 200 Gbps per VM Tier_1 networking performance. C3D is powered by fourth generation AMD EPYC Genoa processors and offers a sustained, all-core turbo frequency of 3.3 GHz, 2,880 GB of DDR5 memory, and up to 200 Gbps per VM Tier_1 networking performance.

For bare metal machine types, choose the C3 machine series.

All third and fourth generation general-purpose VMs supportTitanium.

E2, E2 shared-core, N2, N2D, Tau T2A, and Tau T2D are second generation machine series in this family; N1 and its related shared-core machine types are the first generation machine series.

To learn how your selection affects the performance of Persistent Disk volumes attached to your VMs, seeConfigure your Persistent Disk and VMs.

C4D machine series

C4D VMs are powered by the fifth generation AMD EPYC Turin processor andTitanium. C4D delivers a 30% performance boost over C3D on the estimatedSPECrate®2017_int_base benchmark, which lets you scale performance with fewer resources, thereby optimizing your costs.

C4D is designed to run workloads including web, app and game servers, AI inference, web serving, video streaming, and data centric applications like analytics, as well as relational and in-memory databases.

For databases, C4D delivers 56% more queries per second on Cloud SQL for MySQL and 38% higher operations per second for Memorystore for Redis workloads compared to C3D due to its higher core frequency (up to 4.1 GHz) and improved Instructions Per Clock (IPC).

For web-serving workloads, AMD EPYC Turin's advancements in L3-cache efficiency and branch prediction enable up to 80% higher throughput per vCPU with C4D.

In summary, the C4D machine series has the following features:

• Powered by the AMD EPYC Turin CPU and Titanium
• Offers predefined machine types that range in size from 2 to 384 vCPUs
• Supports up to 3,024 GB of DDR5 memory for VM instances and up to 3,072 GB of memory for bare metal instances
• Supports up to 12,000 GiB of local Titanium SSD disks
• Supports consumption options like on-demand, Spot VMs, and future reservations
• Supports standard network configuration with up to 100 Gbps bandwidth
• Supports per VM Tier_1 networking performance with up to 200 Gbps bandwidth
• Supports only Hyperdisk volumes
• Supports Confidential VM with AMD SEV
• Supports Resource-based and flexible committed use discounts
• Supports compact and spread placement policies

C4D machine types

C4D VMs are available as predefined configurations in standard, highcpu, and highmem sizes ranging from 2 vCPU to 384 vCPUs and up to 3,024 GB of memory.

To use Titanium SSD with C4D, create your instance using the -lssd variant of the C4D machine types. Selecting this machine type creates an instance of the specified size with Titanium SSD partitions attached. You can't attach Titanium SSD volumes separately.

To create a bare metal instance with C4D, use one of the following machine types:

• c4d-standard-384-metal
• c4d-highcpu-384-metal
• c4d-highmem-384-metal

C4D

standard

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4D

highcpu

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4D

highmem

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4D standard

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4D highmem

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4D doesn't support custom machine types.

Regional availability for C4D instances

For C4D VMs, you can view the available regions and zones in theAvailable regions and zones table, as follows:

• To view all the zones where you can create a C4D VM, in theSelect a machine series menu, select C4D.
• You can also use the Select a locationmenu to limit the results to a geographical area.

For regional availability of C4D bare metal instances, seeBare metal instances on Compute Engine.

Supported disk types for C4D

C4D VMs support only the NVMe disk interface and can use the followingHyperdisk block storage:

• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Extreme (hyperdisk-extreme)
• Local Titanium SSD(added automatically with -lssd machine types)

C4D doesn't support Persistent Disk.

Disk and capacity limits

You can use a mixture of different Hyperdisk types with a VM, but the maximum total disk capacity (in TiB) across all disk types can't exceed:

• For machine types with less than 32 vCPUs: 257 TiB for all Hyperdisk
• For machine types with 32 or more vCPUs: 512 TiB for all Hyperdisk

For details about the capacity limits, see Hyperdisk size and attachment limits.

C4D storage limits are described in the following table:

C4D standard

C4D highcpu

C4D highmem

C4D standard Local SSD

C4D highmem with Local SSD

Network support for C4D instances

The following network interface drivers are required:

• C4D VM instances requiregVNIC network interfaces.
• C4D bare metal instances require theIntel IDPF LAN PF device driver.

C4D supports up to 100 Gbps network bandwidth for standard networking and up to 200 Gbps with per VM Tier_1 networking performance for VM and bare metal instances.

Before migrating to C4D or creating C4D VMs or bare metal instances, make sure that theoperating system imagethat you use supports the IDPF network driver for bare metal instances or the gVNIC driver for VM instances. To get the best possible performance on C4D VMs, choose an OS image that supports both "Tier_1 Networking" and "200 Gbps network bandwidth". These images include an updated gVNIC driver, even if the guest OS shows the gve driver version as 1.0.0. If your C4D VM is using an operating system with an older version of gVNIC driver, this is still supported but the VM might experience suboptimal performance such as less network bandwidth or higher latency.

If you use a custom OS image to create a C4D VM, you canmanually install the most recent gVNIC driver. The gVNIC driver version v1.3.0 or later is recommended for use with C4D VMs. Google recommends using the latest gVNIC driver version to benefit from additional features and bug fixes.

C4A machine series

C4A VMs are powered by Google's first Arm-based Axion™ processor. C4A provides machine types with up to 72 vCPUs and 576 GB of DDR5 memory, and 6 TiB of local Titanium SSD. C4A is available in standard, highmem, and highcpu machine types, and also offers-lssd variants for Titanium SSD. C4A uses Google Cloud's latest generation of storage options including Hyperdisk Balanced, Hyperdisk Extreme, and Titanium SSD. C4A offers up to 50 Gbps of standard network performance, and up to 100 Gbps per VM Tier_1 networking performance for your instances.

C4A VMs are placed within a single node withUniform Memory Access (UMA)and also support sole tenant nodes to deliver consistent performance.

In summary, the C4A machine series has the following features:

• Is powered by the Google Axion CPU and Titanium.
• Supports up to 72 vCPUs and 576 GB of DDR5 memory.
• Supports up to 6 TiB of local Titanium SSD disks.
• Offers multiple predefined machine types.
• Supports standard network configuration with up to 50 Gbps bandwidth.
• Supports per VM Tier_1 networking performance with up to 100 Gbps bandwidth.
• Supports Hyperdisk only.
• Supports the following discount and consumption options:
  • Resource-based and flexible committed use discounts
  • Spot VMs
  • Reservations
• Supports the performance monitoring unit (PMU).
• Doesn't support compact placement policies.
• Doesn't support suspendwith C4A instances that have attached Titanium SSD disks.

For information about migrating to Arm VMs, read theArm on Compute document.

C4A machine types

C4A VMs are available as predefined configurations in sizes ranging from 1 vCPU to 72 vCPUs and up to 576 GB of memory.

• standard: 4 GB memory per vCPU
• highcpu: 2 GB memory per vCPU
• highmem: 8 GB memory per vCPU

To use Titanium SSD with C4A, create your VM using the -lssd variant of the C4A machine types. Selecting this machine type creates a VM of the specified size with Titanium SSD partitions attached. You can't attach Titanium SSD volumes separately.

C4A

standard

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C4A

highcpu

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C4A

highmem

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C4A standard

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C4A highmem

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C4A doesn't support custom machine types.

Supported disk types for C4A

C4A VMs support only the NVMe disk interface and can use the followingHyperdisk block storage:

• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Balanced High Availability (hyperdisk-balanced-high-availability)
• Hyperdisk Extreme (hyperdisk-extreme)
• Local Titanium SSD(added automatically with -lssd machine types)

C4A doesn't support Persistent Disk.

Disk and capacity limits

You can use a mixture of different Hyperdisk types with a VM, but the maximum total disk capacity (in TiB) across all disk types can't exceed:

• For machine types with less than 32 vCPUs: 257 TiB for all Hyperdisk
• For machine types with 32 or more vCPUs: 512 TiB for all Hyperdisk

For details about the capacity limits, see Hyperdisk size and attachment limits.

C4A standard

C4A highcpu

C4A highmem

C4A standard

C4A highmem

Network support for C4A instances

C4A instances requiregVNIC network interfaces. C4A instances support up to 50 Gbps network bandwidth for standard networking and up to 100 Gbps network bandwidth per VM Tier_1 networking performance.

Before migrating to C4A or creating C4A instances, make sure that the operating system image that you use supports the gVNIC driver. To get the best possible performance on C4A instances, on the Networking featurestab of the OS details table, choose an OS image that supports both "Tier_1 Networking" and "200 Gbps network bandwidth". These images include an updated gVNIC driver, even if the guest OS shows the gve driver version as 1.0.0. If your C4A instance is using an operating system with an older version of the gVNIC driver, this is still supported but the instance might experience suboptimal performance such as less network bandwidth or higher latency.

If you use a custom OS image with the C4A machine series, you canmanually install the most recent gVNIC driver. The gVNIC driver version v1.3.0 or later is recommended for use with C4A instances. Google recommends using the latest gVNIC driver version to benefit from additional features and bug fixes.

Maintenance experience for C4A instances

During the lifespan of a virtual machine (VM) instance, the host machine that your instance runs undergoes multiple host events. A host event can include the regular maintenance of Compute Engine infrastructure, or in rare cases, a host error. Compute Engine also applies some non-disruptive lightweight upgrades for the hypervisor and network in the background.

The C4A machine series offers the following features related to host maintenance:

The maintenance frequencies shown in the previous table are approximations, not guarantees. Compute Engine might occasionally perform maintenance more frequently.

C4 machine series

C4 VMs are powered by 5th generation Intel Xeon Scalable processors (code-named Emerald Rapids) andTitanium. The C4 machine series is designed to deliver price-performance and enterprise-grade reliability along with a maintenance experience for your most demanding workloads.

Appropriate workloads include web and app serving, game servers, databases and caches, video streaming, data analytics, network appliances, and CPU-based ML inference.

C4 VMs are designed to achieve maximum performance from single-core turbo boosting. You can disable vCPU boosting and limit the vCPUs to the long term sustainable all-core-turbo frequency for more consistent vCPU performance. You can do this by setting turboMode=ALL_CORE_MAX in theAdvancedMachineFeaturessettings.

In summary, the C4 machine series:

• Is powered by 5th generation Intel Emerald Rapids processor and Titanium processors
• Lets you switch between core-boosting performance and steady all-core turbo performance for your vCPUs.
• Supports up to 192 vCPUs and 1.5 TB of DDR5 memory
• Offers multiple predefined machine types
• Supports standard network configuration with up to 100 Gbps bandwidth
• Supports per VM Tier_1 networking performance with up to 200 Gbps bandwidth
• Supports Intel Advanced Matrix Extensions (AMX), a built-in accelerator that significantly improves the performance of deep-learning training and inference on the CPU.
• Supports the following discount and consumption options:
  • Resource-based and flexible committed use discounts
  • Spot VMs
  • Reservations
• Supports the performance monitoring unit (PMU).

C4 machine types

C4 VMs are available as predefined configurations in sizes ranging from 2 vCPUs to 192 vCPUs and up to 1,488 GB of memory.

• standard: 3.75 GB memory per vCPU
• highcpu: 2 GB memory per vCPU
• highmem: 7.75 GB memory per vCPU

C4 standard

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4 highcpu

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4 highmem

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networkingfor larger machine types.

C4 doesn't support custom machine types.

Supported disk types for C4

C4 VMs support only the NVMe disk interface and can use the followingHyperdisk block storage:

• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Extreme (hyperdisk-extreme)

C4 does not support Persistent Disk or Local SSD. ReadMove your workload from an existing VM to a new VMto migrate your Persistent Disk resources to a newer machine series.

Disk and capacity limits

You can use a mixture of different Hyperdisk types with a VM, but the maximum total disk capacity (in TiB) across all disk types can't exceed:

• For machine types with less than 32 vCPUs: 257 TiB for all Hyperdisk
• For machine types with 32 or more vCPUs: 512 TiB for all Hyperdisk

For details about the capacity limits, see Hyperdisk size and attachment limits.

The number of Hyperdisk volumes of all types that you can attach to a VM can't exceed the limits stated in the Max number of Hyperdisk volumes. For details about these limits, seeHyperdisk size and attachment limits.

The combined total number of Hyperdisk Balanced and Hyperdisk Extreme volumes attached to a single VM depends on the number of vCPUs the VM has, and can't exceed these limits:

C4 standard

C4 highcpu

C4 highmem

Network support for C4 VMs

C4 VM instances requiregVNIC network interfaces. C4 supports up to 100 Gbps network bandwidth for standard networking and up to 200 Gbps with per VM Tier_1 networking performance.

Before migrating to C4 or creating C4 instances, make sure that the operating system image that you use supports the gVNIC driver. To get the best possible performance on C4 instances, on the Networking featurestab of the OS details table, choose an OS image that supports both "Tier_1 Networking" and "200 Gbps network bandwidth". These images include an updated gVNIC driver, even if the guest OS shows the gve driver version as 1.0.0. If your C4 instance is using an operating system with an older version of the gVNIC driver, this is still supported but the instance might experience suboptimal performance such as less network bandwidth or higher latency.

If you use a custom OS image with the C4 machine series, you canmanually install the most recent gVNIC driver. The gVNIC driver version v1.3.0 or later is recommended for use with C4 instances. Google recommends using the latest gVNIC driver version to benefit from additional features and bug fixes.

N4 machine series

N4 VMs are powered by the 5th generation Intel Xeon Scalable processors (code-named Emerald Rapids) andTitanium. N4 machine types are built from the ground up for flexibility and cost optimization through an efficient architecture of streamlined features, shapes, and next generationdynamic resource management, which makes better use of resources on host machines. N4 offers flexible options like custom machine types that lets you use choose varied combinations of compute and memory to optimize costs and reduce resource waste. N4 is suited for a variety of general-purpose workloads that don't require peak processing power at all times.

In summary, the N4 machine series:

• Is powered by 5th generation Intel Emerald Rapids processor andTitanium processors.
• Supports up to 80 vCPUs and 640 GB of DDR5 memory.
• Offers multiple predefined machine types andcustom machine types and extended custom memory up to 640 GB.
• Supports standard network configuration with up to 50 Gbps bandwidth
• Supports Intel Advanced Matrix Extensions (AMX), a built-in accelerator that significantly improves the performance of deep-learning training and inference on the CPU.
• Supports the following discount and consumption options:
  • Resource-based and flexible committed use discounts
  • Spot VMs
  • Reservations
• Does not support Local SSD or per VM Tier_1 networking performance.

N4 machine types

N4 VMs are available as predefined configurations in sizes ranging from 2 vCPUs to 80 vCPUs and up to 640 GB of memory.

• standard: 4 GB memory per vCPU
• highcpu: 2 GB memory per vCPU
• highmem: 8 GB memory per vCPU

N4 standard

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms

.

N4 highcpu

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms

.

N4 highmem

* A CPU uses two threads per core, and a vCPU respresents a single thread. See CPU platforms

.

Supported disk types for N4

N4 VMs supports only the NVMe disk interface and can use the followingHyperdisk block storage:

• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Balanced High Availability (hyperdisk-balanced-high-availability)

N4 does not support Persistent Disk or Local SSD. ReadMove your workload from an existing VM to a new VMto migrate your Persistent Disk resources to a newer machine series.

Disk and capacity limits

The number of Hyperdisk volumes of all types that you can attach to a VM can't exceed the limits stated in the Max number of Hyperdisk volumes. For details about these limits, seeHyperdisk capacity.

The combined total number of Hyperdisk Balanced volumes attached to a single VM depends on the number of vCPUs the VM has, and can't exceed these limits:

For VMs running Microsoft Windows and using the NVMe disk interface, the combined number of both Hyperdisk and Persistent Disk attached volumes can't exceed a total of 16 disks. See Known issues. Local SSD volumes are excluded from this issue.

N4 storage limits are described in the following table:

N4 standard

N4 highcpu

N4 highmem

Network support for N4 VMs

N4 instances requiregVNIC network interfaces. N4 instances support up to 50 Gbps network bandwidth for standard networking and don't support per VM Tier_1 networking performance.

Before migrating to N4 or creating N4 instances, make sure that the operating system image that you use supports the gVNIC driver. To get the best possible performance on N4 instances, on the Networking featurestab of the OS details table, choose an OS image that supports both "Tier_1 Networking" and "200 Gbps network bandwidth". These images include an updated gVNIC driver, even if the guest OS shows the gve driver version as 1.0.0. If your N4 instance is using an operating system with an older version of the gVNIC driver, this is still supported but the instance might experience suboptimal performance such as less network bandwidth or higher latency.

If you use a custom OS image with the N4 machine series, you canmanually install the most recent gVNIC driver. The gVNIC driver version v1.3.0 or later is recommended for use with N4 instances. Google recommends using the latest gVNIC driver version to benefit from additional features and bug fixes.

C3D machine series

C3D VMs are powered by the 4th generation AMD EPYC™ (Genoa) processor with a maximum frequency of 3.7 Ghz. C3D machine types are optimized for the underlying hardware architecture to deliver optimal, reliable, and consistent performance.

C3D uses Titanium, which enables higher levels of networking performance, isolation and security. The C3D machine series supports Tier_1 networking bandwidth of up to 100 Gbps and up to 200 Gbps.

In summary, the C3D machine series:

• Is powered by 4th generation AMD EPYC™ processor and Titanium.
• Supports up to 360 vCPUs and 2,880 GB of DDR5 memory.
• Supports standard network configuration with up to 100 Gbps bandwidth and Tier_1 networking with up to 200 Gbps bandwidth.
• Supports the following discount and consumption options:
  • Resource-based committed use discounts (CUDs)
  • Spot VMs
  • Reservations
• Supports Confidential VMwith AMD SEV.

C3D machine types

C3D VMs are available in standard, highcpu, highmem, and lssdconfigurations in sizes ranging from 4 to 360 vCPUs and up to 2,880 GB of memory. The highcpu configuration offers the lowest price per performance for compute-bound workloads that don't require large amounts of memory.

C3D standard

* A CPU uses two threads per core, and a vCPU represents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C3D highcpu

* A CPU uses two threads per core, and a vCPU represents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C3D highmem

* A CPU uses two threads per core, and a vCPU represents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C3D standard with Local SSD

* A CPU uses two threads per core, and a vCPU represents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C3D highmem with Local SSD

* A CPU uses two threads per core, and a vCPU represents a single thread. See CPU platforms.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

C3D doesn't support custom machine types.

Supported disk types for C3D

C3D VMs support only the NVMe disk interface and can use the following block storage types:

• Balanced Persistent Disk (pd-balanced)
• SSD (performance) Persistent Disk (pd-ssd)
• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Balanced High Availability (hyperdisk-balanced-high-availability)
• Hyperdisk ML (hyperdisk-ml)
• Hyperdisk Extreme (hyperdisk-extreme)
• Hyperdisk Throughput (hyperdisk-throughput)
• Local SSD (only available with -lssd machine types)

To use Local SSD with C3D, create your VM using the -lssd variant of the C3D machine types. Selecting this machine type creates a VM of the specified size with Local SSD partitions attached. You must use a machine type that ends in -lssd to use Local SSD with your C3D VM; you can't attach Local SSD volumes separately.

Disk and capacity limits

For VMs running Microsoft Windows and using the NVMe disk interface, the combined number of both Hyperdisk and Persistent Disk attached volumes can't exceed a total of 16 disks. See Known issues. Local SSD volumes are excluded from this issue.

C3D storage limits are described in the following table:

C3D standard

C3D highcpu

C3D highmem

C3D standard with Local SSD

C3D highmem with Local SSD

Network support for C3D VMs

C3D VM instances requiregVNIC network interfaces. C3D supports up to 100 Gbps network bandwidth for standard networking and up to 200 Gbps with per VM Tier_1 networking performance.

Before migrating to C3D or creating C3D instances, make sure that the operating system image that you use supports the gVNIC driver. To get the best possible performance on C3D instances, on the Networking featurestab of the OS details table, choose an OS image that supports both "Tier_1 Networking" and "200 Gbps network bandwidth". These images include an updated gVNIC driver, even if the guest OS shows the gve driver version as 1.0.0. If your C3D instance is using an operating system with an older version of the gVNIC driver, this is still supported but the instance might experience suboptimal performance such as less network bandwidth or higher latency.

If you use a custom OS image with the C3D machine series, you canmanually install the most recent gVNIC driver. The gVNIC driver version v1.3.0 or later is recommended for use with C3D instances. Google recommends using the latest gVNIC driver version to benefit from additional features and bug fixes.

C3 machine series

C3 VMs are powered by the 4th generation Intel Xeon Scalable processors (code-named Sapphire Rapids), DDR5 memory, and Titanium. C3 machine types are optimized for the underlying NUMA architecture to deliver optimal, reliable, and consistent performance.

The new C3 machine series is a major leap in our purpose-built infrastructure offerings:

• Leveraging Titanium processors to offload networking from the CPUs
• Delivering high performance block-storage with Google Cloud Hyperdisk
• Speeding up ML training and inference with Intel AMX

C3 uses Titanium to enable higher levels of networking performance, isolation and security. The C3 machine series supports a default network bandwidth of up to 100 Gbps and up to 200 Gbps withper VM Tier_1 networking performance. Titanium has been designed from the ground up to enable updates that don't impact running workloads.

The C3 machine series provides some of the largest general-purpose machine types, letting you create VM instances with up to 176 vCPUs and 1.4 TB of memory.

C3 has bare metal machine types, which allow you to access all the raw compute resources of the server. You can create bare metal instances with 192 vCPUs and up to 1,536 GB of memory. Bare metal instances also provide access to several onboard, function-specificaccelerators and offloads:

• Intel-QAT
• Intel-DLB
• Intel DSA
• Intel IAA

If your organization uses a Shielded VM policy, then you must create a custom org policy that excludes bare metal shapes before you can create a bare metal instance.

In summary, the C3 machine series:

• Is powered by Intel 4th Generation Xeon processors and Titanium.
• Supports up to 176 vCPUs and 1.4 TB of DDR5 memory for VMs.
• Supports up to 192 vCPUs and 1,536 GB of memory for bare metal instances.
• Supports standard network configuration with up to 100 Gbps bandwidth and Tier_1 networking with up to 200 Gbps bandwidth.
• Supports Intel Advanced Matrix Extensions (AMX), a built-in accelerator that significantly improves the performance of deep-learning training and inference on the CPU.
• Supports the following discount and consumption options:
  • Resource-based committed use discounts (CUDs)
  • Flexible CUDs
  • Spot VMs
  • Reservations
• SupportsConfidential VMwith Intel TDX.
• Doesn't support the following:
  • Sustained use discounts (SUDs)
• C3 bare metal instances don't support the following:
  • Google Kubernetes Engine
  • Shielded VM
  • Nested virtualization

C3 machine types

C3 VMs are available in predefined machine types with sizes ranging from 4 to 176 vCPUs and up to 1,408 GB of memory.

To use Local SSD with C3, create your VM using the -lssd variant of the C3 machine types. Selecting this machine type creates a VM of the specified size with Local SSD partitions attached. You must use a c3-standard-*-lssdmachine type to use Local SSD with your VM; you can't attach Local SSD volumes separately.

To create a bare metal instance with C3, use one of the following machine types:

• c3-standard-192-metal
• c3-highcpu-192-metal
• c3-highmem-192-metal

C3 standard

* A vCPU represents a single hardware thread, or logical core.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

C3 highcpu

* A vCPU represents a single hardware thread, or logical core.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

C3 highmem

* A vCPU represents a single hardware thread, or logical core.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

C3 with Local SSD

* A vCPU represents a single hardware thread, or logical core.
† For bare metal instances, the number of vCPUs is equivalent to the number of hardware threads on the host server.
‡ Default egress bandwidth can't exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

C3 doesn't support custom machine types.

C3 regional availability for bare metal instances

For C3 VMs, you can view the available regions and zones in theAvailable regions and zones table.

• Select C3 in the Select a machine type drop-down menu to see all the zones where you can create a C3 VM.
• You can also use the Select a locationdrop-down menu to limit the results to a geographical area.

C3 bare metal instances are available in the following regions and zones:

Supported disk types for C3

C3 VMs support only the NVMe disk interface and can use the following block storage types:

VM instances

• Zonal balanced Persistent Disk (pd-balanced)
• Zonal SSD (performance) Persistent Disk (pd-ssd)
• Hyperdisk Extreme (hyperdisk-extreme)—Requires at least 64 vCPUs
• Hyperdisk ML (hyperdisk-ml)
• Hyperdisk Throughput (hyperdisk-throughput)
• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Balanced High Availability (hyperdisk-balanced-high-availability)
• Local SSD—only available with -lssd machine types

Bare metal instances

• Hyperdisk Balanced (hyperdisk-balanced)
• Hyperdisk Extreme (hyperdisk-extreme)

A set amount of Local SSD disks are added to the C3 VM when you use the -lssdmachine type. This is the only way to include Local SSD storage with a C3 VM. You can't use Local SSD disks with bare metal instances.

Disk and capacity limits

For VMs running Microsoft Windows and using the NVMe disk interface, the combined number of both Hyperdisk and Persistent Disk attached volumes can't exceed a total of 16 disks. See Known issues. Local SSD volumes are excluded from this issue.

C3 storage limits are described in the following table:

C3 standard

C3 highcpu

C3 highmem

C3 with Local SSD

Network support for C3 VMs

The following network interface drivers are required:

• C3 VM instances requiregVNIC network interfaces.
• C3 bare metal instances require theIntel IDPF LAN PF device driver.

C3 supports up to 100 Gbps network bandwidth for standard networking and up to 200 Gbps with per VM Tier_1 networking performance for VM and bare metal instances.

Before migrating to C3 or creating C3 VMs or bare metal instances, make sure that theoperating system imagethat you use supports the IDPF network driver for bare metal instances or the gVNIC driver for VM instances. To get the best possible performance on C3 VMs, choose an OS image that supports both "Tier_1 Networking" and "200 Gbps network bandwidth". These images include an updated gVNIC driver, even if the guest OS shows the gve driver version as 1.0.0. If your C3 VM is using an operating system with an older version of gVNIC driver, this is still supported but the VM might experience suboptimal performance such as less network bandwidth or higher latency.

If you use a custom OS image to create a C3 VM, you canmanually install the most recent gVNIC driver. The gVNIC driver version v1.3.0 or later is recommended for use with C3 VMs. Google recommends using the latest gVNIC driver version to benefit from additional features and bug fixes.

N2D machine series

The N2D machine series can run on either AMD EPYC Milan or AMD EPYC Rome processors. The third generationAMD EPYC Milan processoris available only in specific regions and zones. To use AMD Milan as yourminimum CPU platform, request it when you create your VM instance.

The N2D series provides some of the largest general-purpose machine types with up to 224 vCPUs and 896 GB of memory and vCPU to memory ratios of 1:1, 1:4, and 1:8. The AMD EPYC Rome processors in this series run with a base frequency of 2.25 GHz, an effective frequency of 2.7 GHz, and a max boost frequency of 3.3 GHz.

In summary, the N2D series:

• Support up to 224 vCPUs and 896 GB of memory.
• Support 50 Gbps and 100 Gbpshigh-bandwidth network configurations.
• Available in predefined and custom VMs.
• Offer higher memory-to-core ratios for VMs created with the extended memory feature. Using the extended memory feature helps you avoid per-CPU software licensing costs while providing access to more than 8 GB of memory per vCPU.
• Powered by third generation AMD EPYC Milan and second generation AMD EPYC Rome processors.
• Supports the following discount and consumption options:
  • Resource-based committed use discounts (CUDs)
  • Flexible CUDs
  • Sustained use discounts (SUDs)
  • Spot VMs
  • Reservations
• Does not support GPUs or nested virtualization.
• Supports Confidential VMwith AMD SEV and AMD SEV-SNP.

N2D VMs don't support GPUs or nested virtualization.

N2D machine types

The following table lists the features of the N2D machine series. For some machine types, certain features are not applicable (N/A).

N2D standard

N2D standard machine types have 4 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

N2D high-mem

N2D high-mem machine types have 8 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

N2D high-cpu

N2D high-cpu machine types have 1 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

For details on the pricing information, see the following:

• For machine type pricing, see VM pricing page.
• Disk usage and network usage is charged separately from machine type pricing. For details, see Disk and image pricing and Network pricing.
• For per VM Tier_1 network performance billing rates, seeTier_1 higher bandwidth network pricing.

Supported disk types for N2D

N2D VMs can use the following block storage types:

• Zonal and regional standard Persistent Disk (pd-standard)
• Zonal and regional balanced Persistent Disk (pd-balanced)
• Zonal and regional SSD Persistent Disk (pd-ssd)
• Hyperdisk Throughput (hyperdisk-throughput)
• Local SSD

N2D VMs with Confidential Computing running Microsoft Windows with the NVMe disk interface have a disk attachment limitation of 16 disks. SeeKnown issuesfor details.

N2D standard

* The maximum size per Persistent Disk volume is 64 TiB.

†The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡ The maximum total disk size applies to all Persistent Disk and Hyperdisk disk types attached to the VM.

N2D high-mem

* The maximum size per Persistent Disk volume is 64 TiB.

†The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡ The maximum total disk size applies to all Persistent Disk and Hyperdisk disk types attached to the VM.

N2D high-cpu

* The maximum size per Persistent Disk volume is 64 TiB.

†The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡ The maximum total disk size applies to all Persistent Disk and Hyperdisk disk types attached to the VM.

N2 machine series

The N2 machine series has flexible sizing between 2 to 128 vCPUs and 0.5 to 8 GB of memory per vCPU. Machine types in this series run on the following processors:

• Ice Lake—offered in specificregions and zones. It is the default processor for larger machine types.
• Cascade Lake—the default for machine types up to 80 vCPUs. If you want to create VMs with Ice Lake, you must set it as theminimum CPU platform.

You can find more details about these two processors on theCPU platforms page.

Workloads that can take advantage of the higher clock frequency are a good choice for this series. These workloads can get higher per-thread performance while benefiting from all the flexibility that the general-purpose machine family offers.

In summary, the N2 machine series:

• Supports up to 128 vCPUs and 864 GB of memory.
• Supports 50 Gbps, 75 Gbps, and 100 Gbpshigh-bandwidth network configurations.
• Is available in predefined and custom VMs.
• Has higher memory-to-core ratios for VMs created with the extended memory feature. Using the extended memory feature helps control per-CPU software licensing costs while providing access to more than 8 GB of memory per vCPU.
• Supports the following discount and consumption options:
  • Resource-based committed use discounts (CUDs)
  • Flexible CUDs
  • Sustained use discounts (SUDs)
  • Spot VMs
  • Reservations

N2 machine types

N2 standard

N2 standard machine types have 4 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

N2 high-mem

N2 high-memory machine types have 8 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

N2 high-cpu

N2 high-cpu machine types have 1 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supports high-bandwidth networking for larger machine types.

For details on the pricing information, see the following:

• For machine type pricing, see VM pricing page.
• Disk usage and network usage is charged separately from machine type pricing. For details, see Disk and image pricing and Network pricing.
• For per VM Tier_1 network performance billing rates, seeTier_1 higher bandwidth network pricing.

Supported disk types for N2

N2 VMs can use the following block storage types:

• Zonal and regional standard Persistent Disk (pd-standard)
• Zonal and regional balanced Persistent Disk (pd-balanced)
• Zonal and regional SSD Persistent Disk (pd-ssd)
• Extreme Persistent Disk (pd-extreme)
• Hyperdisk Extreme (hyperdisk-extreme). Not supported with custom N2 machine types.
• Hyperdisk Throughput (hyperdisk-throughput)
• Local SSD

N2 standard

* The maximum size per Persistent Disk volume is 64 TiB.

† The maximum size per Hyperdisk Extreme volume is 64 TiB. The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡ You can attach a mixture of Hyperdisk and Persistent Disk volumes to a VM, but the total Persistent Disk capacity can't exceed 257 TiB.

N2 high-mem

* The maximum size per Persistent Disk volume is 64 TiB.

† The maximum size per Hyperdisk Extreme volume is 64 TiB. The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡ You can attach a mixture of Hyperdisk and Persistent Disk volumes to a VM, but the total Persistent Disk capacity can't exceed 257 TiB.

N2 high-cpu

* The maximum size per Persistent Disk volume is 64 TiB.

† The maximum size per Hyperdisk Extreme volume is 64 TiB. The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡ You can attach a mixture of Hyperdisk and Persistent Disk volumes to a VM, but the total Persistent Disk capacity can't exceed 257 TiB.

E2 machine series

The cost-optimized E2 machine series have between 2 to 32 vCPUs with a ratio of 0.5 GB to 8 GB of memory per vCPU for standard VMs, and 0.25 to 1 vCPUs with 0.5 GB to 8 GB of memory for shared-core E2 machine types. The E2 machine series offers both Intel and AMD EPYC processors. The processor is selected for you at the time of VM creation. Machine types in this series are available in all regions and zones and support a virtio memory balloon device.

In summary, the E2 machine series:

• Supports up to 32 vCPUs and 128 GB of memory.
• Supports Intel and AMD EPYC Rome and Milan processors.
• Is available in predefined and custom VMs.
• Offers the lowest on demand pricing across the general-purpose machine types.
• Supports the following discount and consumption options:
  • Resource-based committed use discounts (CUDs)
  • Flexible CUDs
  • Spot VMs
  • Reservations
• Does not offer sustained use discounts (SUDs); however, it provides consistently low on-demand and committed-use pricing.
• Does not support GPUs, Local SSDs, sole-tenant nodes, or nested virtualization.

Shared-core VMs

The E2 machine series also contains shared-core machine types that use context- switching to time-share a physical core between vCPUs for multitasking. Different shared-core machine types sustain different amounts of time on a physical core.

Unlike predefined machine types and custom machine types, shared-core machine types have a predefined price that includes both vCPUs and memory. For more information, seeVM instance pricing.

CPU bursting

Shared-core machine types offer bursting capabilities that allow instances to use additional physical CPU for short periods of time. Bursting happens automatically when your VM requires more physical CPU than originally allocated. During these spikes, your VM will opportunistically take advantage of available physical CPU in bursts. Note that bursts are not permanent and are only possible periodically.

• e2-micro, e2-small, and e2-medium shared-core VMs can burst for dozens of seconds.
• If the CPU is utilized at 100%, then the burst lasts as follows:
  • e2-micro: 30 seconds
  • e2-small: 60 seconds
  • e2-medium 120 seconds

The exact burst time is determined by aToken bucketmeaning utilizing the CPU less than 100% will result in longer bursts. Bursting doesn't incur any additional charges. You are charged the listed on-demand price for e2 shared-core, f1-micro, and g1-small VMs.

E2 Limitations

• The E2 machine series does not offer SUDs; however, it provides consistently low on-demand and committed-use pricing.
• The E2 machine series does not support GPUs, Local SSDs, sole-tenant nodes, or nested virtualization.

E2 machine types

E2 is available in standard, highmem, and highcpu configurations, as well as shared-core machine type. In general, E2 shared-core machine types can be more cost-effective for running small, non-resource intensive applications than standard, high-memory, or high-CPU machine types.

E2 standard

E2 standard machine types have 4 GB of system memory per vCPU.

† Persistent Disk and Hyperdisk usage is charged separately from machine pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual See Network bandwidth.

E2 high-mem

E2 high-memory machine types have 8 GB of system memory per vCPU.

† Persistent Disk and Hyperdisk usage is charged separately frommachine pricing.
‡Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

E2 high-cpu

E2 high-cpu machine types have 1 GB of system memory per vCPU.

† Persistent Disk and Hyperdisk usage is charged separately from machine pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

E2 shared-core

E2 shared-core machine types are cost-effective, have a virtio memory balloon device, and are ideal for small workloads. When you use E2 shared-core machine types, your VM runs two vCPUs simultaneously shared on one physical core for a specific fraction of time depending on the machine type. Each vCPU can burst up to 100% of CPU time, for short periods, before returning to the time limitations described in the preceding section.

• e2-micro sustains 2 vCPUs, each for 12.5% ofCPU time totaling 25% CPU time.
• e2-small sustains 2 vCPUs, each at 25% ofCPU time totaling 50% CPU time.
• e2-medium sustains 2 vCPUs, each at 50% ofCPU time totaling 100% CPU time.

* Fractional vCPU of 0.25, 0.5, or 1.0 with 2 vCPUs exposed to the guest operating system.
† Persistent Disk and Hyperdisk usage is charged separately from machine pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supported disk types for E2 VMs

E2 VMs can use the following block storage types:

• Zonal and regional balanced Persistent Disk (pd-balanced)
• Zonal and regional SSD Persistent Disk (pd-ssd)
• Zonal and regional standard Persistent Disk (pd-standard)

N1 machine series

The N1 machine series is Compute Engine's first generation general-purpose machine series available on Intel Skylake, Broadwell, Haswell, Sandy Bridge, and Ivy Bridge CPU platforms.

In summary, the N1 machine series offers the following features:

• Supports up to 96 vCPUs and 624 GB of memory.
• Has both predefined machine types and custom machine types. Custom machine types can be created within a wide range of memory-to-core ratio, ranging from 1 GB per vCPU to 6.5 GB per vCPU.
• Offers higher memory-to-core ratios for VMs created with the extended memory feature.
• Supports the following discount and consumption options:
  • Resource-based committed use discounts (CUDs)
  • Flexible CUDs
  • Sustained use discounts (SUDs); N1 machine series offers a higher SUD percentage than the N2 machine series.
  • Spot VMs
  • Reservations
• Supports Tensor Processing Units (TPUs) in selectzones.
• Can support up toten virtual interfaces per instance.

N1 machine types

N1 standard

N1 standard machine types have 3.75 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
† Persistent Disk and Hyperdisk usage is charged separately frommachine type pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

32 Gbps for Skylake or later CPU platforms. 16 Gbps for all other platforms.

N1 high-memory

N1 high-memory machine types have 6.5 GB of system memory per vCPU.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
† Persistent Disk and Hyperdisk usage is charged separately frommachine type pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

32 Gbps for Skylake or later CPU platforms. 16 Gbps for all other platforms.

N1 high-cpu

N1 high-cpu machine types have 0.9 GB of system memory per vCPU. High-CPU machine types are ideal for tasks that require a moderate increase of vCPUs relative to memory.

* A vCPU is implemented as a single hardware thread, or logical core, on one of the availableCPU platforms.
† Persistent Disk and Hyperdisk usage is charged separately frommachine type pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

32 Gbps for Skylake or later CPU platforms. 16 Gbps for all other platforms.

N1 shared-core

Each vCPU can burst up to 100% of CPU time, for short periods, before returning to their prescribed CPU time limitations.

• When you run an f1-micro, your VM sustains a single vCPU for up to 20% of CPU time.
• When you run a g1-small, your VM sustains a single vCPU for up to 50% of CPU time.

* Fractional vCPU of 0.2 or 0.5, with 1 vCPU exposed to the guest operating system.
† Persistent Disk and Hyperdisk usage is charged separately from VM pricing.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Supported disk types for N1 VMs

N1 VMs can use the following block storage types:

• Zonal and regional balanced Persistent Disk (pd-balanced)
• Zonal and regional SSD Persistent Disk (pd-ssd)
• Zonal and regional standard Persistent Disk (pd-standard)
• Local SSD disks

Tau T2A machine series

The Tau T2A machine series runs on the Ampere Altra Arm processor with a base frequency of 3.0 GHz. Tau T2A offers predefined machine types with 1 to 48 vCPUs, supports 4 GB of memory per vCPU, and offers a maximum of 32 Gbps of outbound data transfer.

This series is available only in selectregions and zones.

The Tau T2A machine series doesn't support simultaneous multithreading (SMT); each vCPU is equivalent to an entire core.

Tau T2A machine types

Tau T2A standard machine types have 4 GB of system memory per vCPU.

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

Tau T2A Limitations

The Tau T2A machine series does not support:

• Custom machine types
• Sole tenant nodes
• Nested virtualization
• Extreme Persistent Disk
• Local SSD
• Regional Persistent Disk
• Virtio-SCSI Storage Controller and Virtio-Net Ethernet Adapter
• Windows Server or Windows Client OS
• 32-bit mode EL0 (guest userspace support)
• Committed use discounts (CUDs)or Sustained use discounts (SUDs); however, it offers Spot VM discounts.
• Virtual display devices

T2A supports the Secure bootfeature, but not all public OS images for T2A support secure boot.

Supported disk types for T2A

T2A VMs support only the NVMe disk interface and can use the following block storage types:

• Zonal standard Persistent Disk (pd-standard)
• Zonal balanced Persistent Disk (pd-balanced)
• Zonal SSD (performance) Persistent Disk (pd-ssd)

For VMs running Microsoft Windows and using the NVMe disk interface, the combined number of both Hyperdisk and Persistent Disk attached volumes can't exceed a total of 16 disks. See Known issues. Local SSD volumes are excluded from this issue.

Tau T2D machine series

The Tau T2D machine series run on the third generationAMD EPYC Milan processorwith a base frequency of 2.45 GHz, an effective frequency of 2.8 GHz, and a max boost frequency of 3.5 GHz. This series has predefined machine types of up to 60 vCPUs, support 4 GB of memory per vCPU, and a maximum of 32 Gbps outbound data transfer. It also supports the following discount and consumption options:

• Resource-based committed use discounts (CUDs)
• Spot VMs
• Reservations

This series is available only in selectregions and zones.

Machine types in the Tau T2D machine series have simultaneous multithreading (SMT) disabled; therefore a vCPU is equivalent to an entire core.

Tau T2D Limitations

Tau T2D VMs don't support:

• Local SSD
• Regional Persistent Disk
• Custom VMs
• Sole-tenant nodes
• Extreme Persistent Disk
• GPUs
• Nested virtualization
• Flexible CUDs
• Sustained use discounts (SUDs)
• Confidential VMs

Tau T2D machine types

Tau T2D standard machine types have 4 GB of system memory per vCPU.

* SMT is not supported. Each vCPU is equivalent to an entire core. See CPU platforms.
‡ Maximum egress bandwidth cannot exceed the number given. Actual egress bandwidth depends on the destination IP address and other factors. See Network bandwidth.

For details on the pricing information, see the following:

• For machine type pricing, see VM pricing page.
• Disk usage and network usage is charged separately from machine type pricing. For details, see Disk and image pricing and Network pricing.

Supported disk types for T2D

T2D VMs can use the following block storage types:

• Zonal standard Persistent Disk (pd-standard)
• Zonal balanced Persistent Disk (pd-balanced)
• Zonal SSD (performance) Persistent Disk (pd-ssd)
• Hyperdisk Throughput (hyperdisk-throughput)

* The maximum size per Persistent Disk volume is 64 TiB.

†The maximum size per Hyperdisk Throughput volume is 32 TiB.

‡You can attach a mixture of Hyperdisk and Persistent Disk volumes to a VM, but the total Persistent Disk capacity can't exceed 257 TiB.

Custom machine types

If none of the predefined machine types in the general-purpose machine family match your workload needs, you can create a VM with a custom machine type.

Creating a VM with a custom machine type is ideal for workloads that require more processing power or more memory, but don't need all of the upgrades that are provided by the next larger predefined machine type.

It costs slightly more to use a custom machine type than an equivalent predefined machine type, and there are limitations in the amount of memory and vCPUs that you can select. The on-demand prices for custom machine types include a 5% premium over the on-demand and commitment prices for predefined machine types.

You can create a VM with a custom machine type for only the N and E machine series in the general-purpose machine family. Custom machine types are not available for the C and Tau machine series. Custom machine types are subject to the same Persistent Disk limits as E2, N2, and N1 predefined machine types. The maximum total Persistent Disk size for each VM is 257 TiB and the max number of Persistent Disk is 128. N4 custom machines types are subject to the limitations ofHyperdisk capacitystorage.

If a custom machine type doesn't meet your requirements, it's possible tocustomize the number of visible CPU coreson many machine types. It's also possible toset the number of threads per corefor certain machine types. You can make these changes during VM instance creation, or by editing an existing VM instance. Reducing the number of visible cores might impact the cost of your VMs. Be sure to reviewpricing prior to making any changes.

N4 custom machine types

• For N4 custom machine types, you can create a machine type with 2 to 80 vCPUs with the vCPUs in multiples of 2, and memory between 4 and 640 GB.
• By default, the memory per vCPU that you can select for a custom machine type is determined by the machine series you use. For the N4 machine series, select between 2 GB and 8 GB per vCPU in 256 MB increments. When creating a standard N4 machine type, the minimum memory you can select is 4 GB. Higher amounts of memory are possible by enablingextended memory.

Examples of invalid machine types:

• 2 vCPUs, 0.5 GB of total memory. Invalid because the total memory is less than the minimum 4 GB for an N4 VM.
• 1 vCPU, 8 GB of memory. Invalid because the vCPU count is too small. N4 custom machine types require a minimum of 2 vCPUs.

Examples of valid machine types:

• 36 vCPUs, 72 GB of total memory. Valid because the total number of vCPUs is even and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 2 GB to 8 GB per vCPU.
• 2 vCPUs, 14 GB of total memory. Valid because it has 2 vCPUs, which is the minimum value, and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 2 GB to 8 GB per vCPU.

N2D custom machine types

• The maximum number of vCPUs allowed for a custom machine type is determined by the machine series you choose. For the N2D machine series, which support AMD EPYC Rome and Milan platforms, you can deploy custom machine types with 2 to 96 vCPUs.
• You can create N2D custom machine types with 2, 4, 8, or 16 vCPUs. After 16, you can increment the number of vCPUs by 16, up to 96 vCPUs. The minimum acceptable number of vCPUs is 2.
• By default, the memory per vCPU that you can select for a custom machine type is determined by the machine series you choose. For N2D machine types, select between 0.5 GB and 8.0 GB per vCPU in 256 MB increments. Higher amounts of memory are possible by enablingextended memory.
• N2D custom machine types are available only in selectregions and zones.
• N2D custom machine types support per VM Tier_1 networking performance maximum egress limits of 50 Gbps to 100 Gbps. When enabled:
  • VMs with 48 to 94 vCPUs have a total egress limit of 50 Gbps.
  • VMs with 96 vCPUs have a total egress limit of 100 Gbps.

Examples of invalid machine types:

• 2 vCPUs, 0.4 GB of total memory. Invalid because the total memory is less than the minimum 1 GB for an N2D VM.
• 34 vCPUs, 34 GB of total memory. Invalid because the total number of vCPUs is not divisible by 16.
• 1 vCPU, 1024 MB of memory. Invalid because the vCPU count is too small. N2D custom machine types require a minimum of 2 vCPUs.

Examples of valid machine types:

• 32 vCPUs, 16 GB of total memory. Valid because the total number of vCPUs is even and the total memory is a multiple of 256 MB. The amount of memory per vCPU is 1 GB, which satisfies the minimum requirement. Because the number of vCPUs is larger than 8 vCPUs, the number of vCPUs must be divisible by 16.
• 2 vCPUs, 7 GB of total memory. Valid because it has 2 vCPUs, which is the minimum value, and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 1 GB to 8 GB per vCPU.

N2 custom machine types

• For N2 custom machine types, you can create a machine type with 2 to 80 vCPUs and memory between 1 and 864 GB. For machine types with up to 32 vCPUs, you can select a vCPU count that is a multiple of 2. For machine types with greater than 32 vCPUs, you must select a vCPU count that is a multiple of 4 (for example, 36, 40, 56, or 80).
• You can create N2 custom machine types on different processors:
  • Cascade Lake, the 2nd generation of the Intel Xeon processor. This is the default processor for N2 custom machine types with less than 80 vCPUs.
  • Ice Lake, the 3rd generation of the Intel Xeon processor. Ice Lake processors are available in specificregions and zones.
• By default, the memory per vCPU that you can select for a custom machine type is determined by the machine series you use. For the N2 machine series, select between 0.5 GB and 8.0 GB per vCPU in 256 MB increments. Higher amounts of memory are possible by enablingextended memory.
• N2 custom machine types have an option for a per VM Tier_1 networking performance maximum egress of 50 Gbps to 100 Gbps with a minimum of 30 vCPUs.
  • 32 to 62 vCPUs have a total egress of 50 Gbps
  • 64 to 78 vCPUs have a total egress of 75 Gbps
  • 80 vCPUs have a total egress of 100 Gbps

Examples of invalid machine types:

• 2 vCPUs, 0.5 GB of total memory. Invalid because the total memory is less than the minimum 1 GB for an N2 VM.
• 34 vCPUs, 34 GB of total memory. Invalid because the total number of vCPUs is not divisible by 4.
• 1 vCPU, 1024 MB of memory. Invalid because the vCPU count is too small. N2 custom machine types require a minimum of 2 vCPUs.

Examples of valid machine types:

• 36 vCPUs, 18 GB of total memory. Valid because the total number of vCPUs is even and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 0.5 GB to 8 GB per vCPU. Because the number of vCPUs is larger than 32 vCPUs, the number of vCPUs must be divisible by 4.
• 2 vCPUs, 7 GB of total memory. Valid because it has 2 vCPUs, which is the minimum value, and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 0.5 GB to 8 GB per vCPU.

E2 custom machine types

• E2 custom machine types support predefined platforms with Intel or AMD EPYC processors. You can create E2 custom machine types with vCPUs in multiples of 2, up to 32 vCPUs. The minimum acceptable number of vCPUs for a VM is 2.
• By default, the general-purpose machine series you choose determines the memory per vCPU that you can select for a custom machine type. For E2, the ratio of memory per vCPU is 0.5 GB to 8 GB inclusive. When creating a standard E2 machine type, the minimum memory you can select is 1 GB.
• An exception to the minimum vCPU limitation is to create an e2-standard-2 VM, then customize the visible core to 1 vCPU. The resulting VM is an e2-custom VM. For example, you create an E2 VM using thee2-standard-2 machine type, stop the VM, and edit it by changing the visible core to 1 vCPU with 1.25 GB of memory. As a result, the machine type changes to e2-custom-2-1280. Pricing is described in theCustomize the number of visible CPU cores document.

Examples of invalid machine types:

• 1 vCPU, 1024 MB of memory. Invalid because the vCPU count is too small. E2 custom machine types require a minimum of 2 vCPUs.
• 32 vCPUs, 1 GB of total memory. Invalid because the ratio of vCPUs to memory is incorrect. The acceptable ratio is 0.5 GB of memory to 1 vCPU.

Examples of valid machine types:

• 32 vCPUs, 16 GB of total memory. Valid because the total number of vCPUs is even and the total memory is an acceptable ratio of memory to vCPU.
• 2 vCPUs, 8 GB of total memory. Valid because it has 2 vCPUs, which is the minimum value, and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 0.5 GB to 8 GB per vCPU.

E2 shared-core custom machine types

E2 shared-core machine types support predefined Intel or AMD EPYC processors, which are preselected for you at the time of VM creation. You can create shared-core machine types with a vCPU range of 0.25 to 1 vCPU. The memory range is 1 to 8 GB, with a maximum ratio of 8 GB per vCPU.

You can't customize the number of visible cores on a shared-core E2 VM.

• e2-micro: 0.25 vCPU, 1 to 2 GB of memory
• e2-small: 0.50 vCPU, 1 to 4 GB of memory
• e2-medium: 1 vCPU, 1 to 8 GB of memory

N1 custom machine types

• You can create N1 custom machine types with 1 or more vCPUs. For VMs with more than 1 vCPU, you must increment the number of vCPUs by 2, up to 96 vCPUs for Intel Skylake platform,or up to 64 vCPUs for Intel Broadwell, Haswell, or Ivy Bridge CPU platforms.
• By default, the memory per vCPU that you can select for a custom machine type is determined by the machine series you choose. For N1 machine types, select between 0.9 GB and 6.5 GB per vCPU, inclusive. N1 custom machine types with 1 or 2 vCPUs require a minimum of 1 GB per vCPU. Higher amounts of memory are possible by enabling extended memory.

Examples of invalid machine types:

• 1 vCPU, 0.2 GB of total memory. Invalid because the total memory is less than the minimum 1 GB for an N1 VM.
• 3 vCPU, 1 GB of total memory. Invalid because the number of vCPU cores must be 1 or an even number up to 96.

Examples of valid machine types:

• 32 vCPUs, 29 GB of total memory. Valid because the total number of vCPUs is even and the total memory is a multiple of 256 MB. The total memory is an acceptable ratio of memory to vCPU.
• 1 vCPU, 1 GB of total memory. Valid because it has one vCPU, which is the minimum value, and the total memory is a multiple of 256 MB. The amount of memory per vCPU is also within the acceptable range of 1 GB to 6.5 GB per vCPU.

What's next

• Network bandwidth
• Configuring a VM with a high-bandwidth network
• Virtual machine instances
• VM instance pricing