Layer 4 · Multi-cloud workload orchestration
SkyPilot
An open-source layer that runs the same job on whichever cloud, specialist GPU provider or Kubernetes cluster has capacity.
Evaluated from public documentation, architecture material and source code. No vendor contact.
What it solves
You define the job once. It finds capacity across your providers, prefers cheaper options such as interruptible instances, and relaunches elsewhere when a node is taken away.
Who it suits
Organisations weighing a capacity commitment who want a credible alternative, and teams whose batch or training work could tolerate interruption if recovery were automatic.
What to watch out for
It orchestrates across providers but does not remove the need to hold accounts and quota with each of them. Quota approval is often the real constraint, not the tooling.
Every product here gets one of these. A recommendation without a trade-off is not a recommendation.
Problems this comes up for
Worth comparing against
Kubernetes batch scheduling
L4 · Batch scheduling
Add-ons giving Kubernetes the queueing and gang-scheduling behaviour AI workloads need and plain Kubernetes does not provide.
- Fit
- Organisations already standardised on Kubernetes.
- Catch
- You assemble a scheduler from components rather than buying one.
Run:ai
L4 · GPU pooling and fractional sharing
A Kubernetes-based orchestration layer that pools accelerators across teams and allocates fractions of one to a workload.
- Fit
- Shared clusters where several teams compete for fixed capacity.
- Catch
- Assumes you already run Kubernetes, and run it well.
Wondering whether SkyPilot is the right choice?
Describe your setup and constraints. We will tell you whether it fits, and what else you should be looking at.