Back to Gap Map Infrastructure / ML orchestration

NVIDIA Run:ai

closed / Overall score: n/a

Enterprise platform for AI workloads and GPU orchestration. It pools GPU capacity across public clouds, private clouds, hybrid environments and on-premise data centres, allocates it dynamically including fractionally, and places AI workloads against quotas set per team.

NVIDIA describes the open KAI Scheduler as 'based on NVIDIA Run:ai'; KAI now lives in its own GitHub organization and is scored separately on this map, so the open component is not credited to this record. The product page is dynamic and returned a different digest on two fetches an hour apart, so a re-fetch will read as drift.

Openness

1 high confidence

1.0

license: Proprietary
service: proprietary(NVIDIA commercial platform)
source: closed(the platform is unpublished

NVIDIA Run:ai is a closed commercial platform. NVIDIA separately open-sourced KAI Scheduler, described as 'based on NVIDIA Run:ai,' but that is a distinct published project scored on its own, so it does not lift this platform's source dimension.

https://www.nvidia.com/en-us/software/run-ai/ recorded 2026-09-15
NVIDIA's Run:ai product page: 'The enterprise platform for AI workloads and GPU orchestration', with dynamic resource allocation and GPU pooling across public clouds, private clouds, hybrid environments and on-premise data centres, and the note that the open-source KAI Scheduler is 'based on NVIDIA Run:ai'. Establishes a closed platform with a separately published component.

Adoption

not assessed

No usage figure is published for this service specifically, and it publishes no countable artifact - no package, repository or registry entry - so no level is assigned rather than one being inferred from the parent platform.

https://www.nvidia.com/en-us/software/run-ai/ recorded 2026-09-15
The NVIDIA Run:ai product page, read for a usage disclosure; it publishes percentage efficiency claims but no count of clusters, GPUs or customers.

Capability

5 medium confidence

5.0

NVIDIA Run:ai's own scheduler documentation states that the Scheduler always handles a pod as part of its pod group, applying gang scheduling, minimum replicas and topology consistently across the whole group, including distributed workloads that span multiple nodes. That is a multi-machine set reserved as one allocation by the platform's own scheduler.

https://run-ai-docs.nvidia.com/saas/platform-management/runai-scheduler/scheduling/concepts-and-principles recorded 2026-09-16
NVIDIA Run:ai's own Scheduler documentation: the Scheduler 'always uses the pod group associated with each pod and handles the pod as part of a larger group of pods, while taking into consideration the common characteristics of the pod group, such as gang scheduling, minimum replicas or workers, workload priority class, workload preemptibility policy, topology information', applied 'consistently across the entire pod group'; workloads range to 'distributed workloads using multiple pods, each running on a node'. Establishes gang scheduling by the platform's own scheduler rather than by a separate component.
https://www.nvidia.com/en-us/software/run-ai/ recorded 2026-09-15
NVIDIA's Run:ai page describing dynamic scheduling and orchestration, GPU pooling and fractional inference allocation, and attributing gang-scheduled deployments to Grove with KAI Scheduler.

Verified 2026-09-15