Atlas Gen2 ARM hardware

Parent: Atlas cluster tiers (M30) & major-version upgrade mechanics · Published reference · snapshot 2026-10-15

Depth-first rabbithole dossier for Atlas Gen2 ARM hardware; source-anchored research pack.

These notes link each claim to its source. Some citations identify archived research reports that are no longer hosted on this site. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.

Definitions

[low] For customers creating new clusters, Gen2 is the recommended default infrastructure where available, and Atlas Gen2 is broadly available across AWS regions that support the underlying infrastructure. (verified as of 2026-10-15) Sources: https://www.mongodb.com/company/blog/product-release-announcements/introducing-atlas-gen2-on-aws-m30-dedicated-clusters [source]
[low] The feature described on this page is not currently supported on Atlas Infinite clusters during public preview. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/reference/google-gcp.md [source]

How it works

[low] When creating a new cluster through the Atlas UI, Atlas automatically selects the Gen2 option for eligible M30+ clusters, but you can select Gen1 to deploy a Gen1 cluster on Atlas Core. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]

Measurements and reference values

[high] Atlas Gen2 on AWS and GCP runs on newer ARM-based processors; Google's C4A instances are powered by Google's Axion processor built on the Arm Neoverse V2 compute core. (The cited sources do not name m8g for AWS or tie C4A to Atlas Gen2 on GCP; those mappings come from the Atlas AWS/GCP reference pages.) Sources: https://www.mongodb.com/company/blog/product-release-announcements/introducing-atlas-gen2-on-aws-m30-dedicated-clusters, https://docs.cloud.google.com/compute/docs/general-purpose-machines [source]
[low] The preferred instance families for Gen2 clusters on AWS are m8g.large (M30, 2 vCPU, 8 GiB), m8g.xlarge (M40, 4 vCPU, 16 GiB), m8g.2xlarge (M50, 8 vCPU, 32 GiB), m8g.4xlarge (M60, 16 vCPU, 64 GiB), m8g.8xlarge (M80, 32 vCPU, 128 GiB), m8g.12xlarge (M140, 48 vCPU, 192 GiB), m8g.16xlarge (M200, 64 vCPU, 256 GiB), and m8g.24xlarge (M300, 96 vCPU, 384 GiB). (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md [source]
[low] The preferred instance families for Gen2 clusters on Google Cloud are c4a-standard-2 (M30, 2 vCPU, 8 GB), c4a-standard-4 (M40, 4 vCPU, 16 GB), c4a-standard-8 (M50, 8 vCPU, 32 GB), c4a-standard-16 (M60, 16 vCPU, 64 GB), c4a-standard-32 (M80, 32 vCPU, 128 GB), c4a-standard-48 (M140, 48 vCPU, 192 GB), and c4a-standard-64 (M200, 64 vCPU, 256 GB). (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/reference/google-gcp.md [source]
[low] In an internal pilot evaluating Atlas clusters migrated to Gen2 infrastructure, MongoDB observed a statistically significant 15–20% reduction in CPU utilization, improved memory efficiency, and approximately 45% lower read and write latency. Sources: https://www.mongodb.com/company/blog/product-release-announcements/introducing-atlas-gen2-on-aws-m30-dedicated-clusters [source]

Problems, failure modes and limitations

[low] All nodes within a cluster must be of the same generation, and you cannot mix Gen1 and Gen2 nodes within the same cluster. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]

Comparisons and alternatives

[low] On Atlas Core clusters, Gen2 Dedicated clusters offer improved performance compared to Gen1 and support Extended Standard IOPS, providing up to 80k IOPS for AWS and 160k IOPS for Google Cloud. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[medium] M30+ Atlas Infinite clusters are available on Gen2 only and do not support Extended Standard IOPS, with storage performance fixed per cluster tier rather than scaling with capacity. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md, https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[medium] Atlas clusters deployed in regions with at least three availability zones are split across three zones, and 3Z clusters have higher availability compared to 2Z clusters, though not all regions support 3Z clusters. Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md, https://www.mongodb.com/docs/atlas/reference/google-gcp.md [source]

Facts and statements

[medium] Atlas supports configuring M30+ Dedicated clusters deployed on AWS and Google Cloud as Gen2 clusters, which are available only on these providers and not on Azure. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md, https://www.mongodb.com/docs/atlas/reference/google-gcp.md, https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[low] Gen2 clusters use Hyperdisk storage on Google Cloud, while Gen1, M10, and M20 clusters use Persistent Disk storage. Sources: https://www.mongodb.com/docs/atlas/reference/google-gcp.md [source]
[low] Gen2 clusters support only reactive auto-scaling, not predictive auto-scaling, and multi-cloud support is not available for Gen2 Dedicated clusters. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[low] For existing clusters, you can switch between Gen1 and Gen2 generations in the same way you would change the cluster's tier. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[low] Not all cloud provider regions support Gen2 Dedicated clusters, and cross-region support is available only if all regions in which you deploy your cluster support Gen2 clusters on your chosen cloud provider. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[low] Atlas Gen2 for M30+ Dedicated Clusters on AWS is available in supported regions across APAC and Europe, while on GCP it is available in supported regions across APAC and Europe with limited availability in select U.S. regions. (verified as of 2026-10-15) Sources: https://www.mongodb.com/company/blog/product-release-announcements/introducing-atlas-gen2-on-aws-m30-dedicated-clusters [source]
[low] M30+ Atlas Core clusters are available in Gen1 or Gen2 on AWS and Google Cloud, while M10 and M20 clusters are generation-agnostic and do not require selecting a cluster generation. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[low] In the API and CLI, Gen1 clusters use only the cluster tier in their names, while Gen2 clusters use the cluster tier with _GEN_2 appended, such as M200_GEN_2. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[low] If you specify a Gen1 value for an existing Gen2 cluster, you might unintentionally change the cluster's generation, so you must specify the corresponding _GEN_2 instance size to keep the cluster on Gen2. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md [source]
[medium] On Atlas Core clusters, storage and Standard IOPS scale independently, so billing reflects only the scaled resources, whereas on Atlas Infinite clusters storage performance is fixed per tier. Sources: https://www.mongodb.com/docs/atlas/manage-clusters.md, https://www.mongodb.com/docs/atlas/reference/amazon-aws.md [source]
[low] Atlas supports all AWS regions other than some regions in China and US GovCloud, while MongoDB Atlas for Government supports US GovCloud regions. Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md [source]
[medium] MongoDB Atlas clusters are always made of replica sets with a minimum of three MongoDB nodes, and the number of zones in a region has no effect on the number of MongoDB nodes Atlas can deploy. Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md, https://www.mongodb.com/docs/atlas/reference/google-gcp.md [source]
[low] The maximum IOPS rate listed for a given instance type in Amazon AWS Documentation may be higher than the IOPS rate Atlas allows you to configure. Sources: https://www.mongodb.com/docs/atlas/reference/amazon-aws.md [source]

Related concepts

Atlas — is a part of Atlas Gen2 ARM hardware
Gen2 — is a part of Atlas Gen2 ARM hardware
ARM — is a part of Atlas Gen2 ARM hardware

Children

No children recorded.

← the whole tree · 3D view· how to read this page