---
type: "post"
id: "modal-clusters-are-generally-available-analysis-00ef08"
slug: "modal-clusters-are-generally-available-analysis-00ef08"
title: "Modal Clusters are generally available"
description: "Thesis: Modal’s release makes multi-node GPU clusters a first-class managed workload in its serverless platform."
retrieval_nugget: "Thesis: Modal’s release makes multi-node GPU clusters a first-class managed workload in its serverless platform."
published_at: "2026-10-01"
updated_at: "2026-10-01"
record_date: "2026-10-01"
date_kind: "published_at"
topics: []
source_urls: ["https://modal.com/blog/modal-clusters-generally-available"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:modal-clusters-are-generally-available-analysis-00ef08"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:646d76df86cfa3b70b740a86df20bca59bee630265e7c30d439e1ab1d2ca5f2e","reviewed_at":"2026-10-01T20:05:19.305Z","source_evidence_count":1,"verified_claim_count":3,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":3,"implication_count":2,"watch_condition_count":2,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Modal’s release makes multi-node GPU clusters a first-class managed workload in its serverless platform.","observed_facts":[{"text":"Modal says Clusters are generally available to all workspaces and launched with the @modal.clustered decorator; its example configures four B300 nodes with rdma=True.","source_urls":["https://modal.com/blog/modal-clusters-generally-available"]},{"text":"The company says its gang scheduler plans placement for all nodes in a cluster together and starts capacity when existing capacity is insufficient.","source_urls":["https://modal.com/blog/modal-clusters-generally-available"]},{"text":"Modal says RDMA is configured for PyTorch and NCCL and can deliver up to 6.4 Tbps.","source_urls":["https://modal.com/blog/modal-clusters-generally-available"]}],"mechanism":"Modal replaces per-node scheduling for these jobs with a gang scheduler that observes fleet state, plans placement for all nodes in a cluster together, and starts capacity when existing capacity is insufficient. It also says RDMA setup for PyTorch and NCCL is automated behind a flag.","why_now":"Modal presents this as a transition from a 1.5-year period of testing to general availability, making the capability accessible across workspaces rather than a limited trial.","implications":["The first implication is a reduction in the operational work needed to request coordinated multi-node capacity, not proof of lower total cost or faster end-to-end training.","Multi-node training and inference can be requested through the same platform primitives as other Modal workloads, potentially reducing infrastructure assembly work for teams."],"evidence_boundary":"The performance, speed, and reliability claims in the announcement are Modal’s own; this source provides no independent benchmark or customer-side verification.","watch_conditions":["Independent workload measurements and customer reports will show whether the reported RDMA throughput and rapid cluster acquisition hold across real deployments.","The next evidence to watch is how cluster scheduling behaves under demand, and whether the announced per-second billing produces predictable costs for real multi-node workloads."],"related_records":[],"new_branch_reason":"No prior New Runtime record is supplied, so this release begins a distinct analytical record about Modal’s general availability of managed multi-node clusters."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/modal-clusters-are-generally-available-analysis-00ef08/","markdown":"https://newruntime.com/posts/modal-clusters-are-generally-available-analysis-00ef08.md","json":"https://newruntime.com/posts/modal-clusters-are-generally-available-analysis-00ef08.json"}
---

# Modal Clusters are generally available

## Retrieval answer

Thesis: Modal’s release makes multi-node GPU clusters a first-class managed workload in its serverless platform.

Thesis: Modal’s release makes multi-node GPU clusters a first-class managed workload in its serverless platform.

Modal says Clusters are generally available to all workspaces and launched with the `@modal.clustered` decorator; its example configures four B300 nodes with `rdma=True`. The company also says clusters integrate with Volumes, Cloud Bucket Mounts, and Queues, and are billed by the second. Mechanism: Modal replaces per-node scheduling for these jobs with a gang scheduler that observes fleet state, plans placement for all nodes in a cluster together, and starts capacity when existing capacity is insufficient. It also says RDMA setup for PyTorch and NCCL is automated behind a flag.

Why now: Modal presents this as a transition from a 1.5-year period of testing to general availability, making the capability accessible across workspaces rather than a limited trial. Implication: Multi-node training and inference can be requested through the same platform primitives as other Modal workloads, potentially reducing infrastructure assembly work for teams. The first implication is a reduction in the operational work needed to request coordinated multi-node capacity, not proof of lower total cost or faster end-to-end training.

Evidence boundary: The performance, speed, and reliability claims in the announcement are Modal’s own; this source provides no independent benchmark or customer-side verification. Watch condition: Independent workload measurements and customer reports will show whether the reported RDMA throughput and rapid cluster acquisition hold across real deployments. Modal says the RDMA fabric can reach up to 6.4 Tbps, but that ceiling alone does not establish workload-level performance.

The next evidence to watch is how cluster scheduling behaves under demand, and whether the announced per-second billing produces predictable costs for real multi-node workloads.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)