6 categories and 32 products and companies, charted by an ai agent.
Cluster schedulers and workload orchestration software allocates compute resources and schedules jobs across GPU clusters. Data centers, ML teams, and enterprises use these systems to manage containerized workloads and batch jobs on distributed hardware.
run:ai · kubernetes · slurm · volcano · kueue · ray · nvidia base command · hpc-ai
Software that schedules and manages GPU workloads across multiple cloud providers and on-premises infrastructure. Data centers, ML teams, and enterprises use it to allocate GPUs to jobs, monitor utilization, and balance load across clouds.
skypilot · dstack · anyscale · determined ai · clearml
Software that schedules, allocates, and manages GPU workloads across multiple cloud providers or on-premise clusters. ML engineers, data scientists, and infrastructure teams use it to provision and distribute compute jobs.
coreweave · lambda · together ai · crusoe · nebius · vast.ai
Serverless GPU platforms provide on-demand GPU compute without infrastructure management, used by ML engineers and data scientists who run training jobs and inference workloads. Users submit containerized code and the platform allocates GPU resources, handles scheduling, and bills by consumption.
modal · runpod · replicate · baseten · beam
Software that allocates, schedules, and manages GPU resources across compute clusters, allowing multiple workloads and users to share physical GPUs through time-slicing, spatial partitioning, or containerization. Used by cloud operators, enterprises, and research institutions running multiple concurrent GPU workloads.
cast ai · exostellar · nvidia dcgm · wekaio
Software that allocates, schedules, and manages GPU resources across clusters and data centers, used by ML engineers and infrastructure teams running training and inference workloads. The systems assign GPUs to jobs, track utilization, and enforce resource limits.
weights & biases · databricks · amazon sagemaker · domino data lab