80+ providers  ·  15+ clouds  ·  UK / Europe / Middle EastBacked by SFC · Rule30 · NVIDIA Inception

Capabilities

Turn your data centre
into a cloud platform

01Managed Kubernetes clusters
02Multi-tenant projects & namespaces
03GPU & CPU scheduling
04Customer quotas & isolation
05Marketplace services
06Observability & billing
07Private networking between sites
08Burst capacity across partners

A Kubernetes distribution designed for infrastructure providers — not just enterprises.

Elasticity

Never lose a customer
to supply constraints

When your GPUs are full, stack8s quietly pulls capacity from partner data centres — delivered white-labelled under your brand on the same Kubernetes control plane your customers already trust.

As your own inventory restocks, workloads migrate back seamlessly — zero downtime, zero customer friction. They never know the difference.

GPU as commodity.
Like electricity.

Source it anywhere — deliver it as yours

01

Supply runs low

Your GPU pool hits capacity — customer demand keeps growing.

02

Burst to partners — white-labelled

stack8s pulls from partner data centres. Same control plane. Your brand. No migration.

03

Restock & reclaim — zero downtime

GPUs back online? Workloads shift home automatically. Customers feel nothing.

Sourced from 36+ partner networks
AWS
Azure
GCP
Oracle
IONOS
Hetzner
OVH
Scaleway
DigitalOcean
Vultr
Contabo
UpCloud
Civo
CoreWeave
Nebius
RunPod
Lambda Labs
Hyperstack
GMI Cloud
Novita
Latitude
Leafcloud
Theta EdgeCloud
FluidStack
Paperspace
Massed Compute
Crusoe
Denvr
Amaya
Boost Run
Horizon Cloud
Verda Cloud
Tencent Cloud
imwt
Voltagepark
wafaicloud
AWS
Azure
GCP
Oracle
IONOS
Hetzner
OVH
Scaleway
DigitalOcean
Vultr
Contabo
UpCloud
Civo
CoreWeave
Nebius
RunPod
Lambda Labs
Hyperstack
GMI Cloud
Novita
Latitude
Leafcloud
Theta EdgeCloud
FluidStack
Paperspace
Massed Compute
Crusoe
Denvr
Amaya
Boost Run
Horizon Cloud
Verda Cloud
Tencent Cloud
imwt
Voltagepark
wafaicloud
Customer retention is easy when they never hit a wall — and never know you hit one either.

Monetisation

Monetise idle
GPU capacity

A provider-side plugin exposes your unused GPU capacity — securely — to customers from other connected data centres. When GPUs aren't being used locally, they're offered through the stack8s control plane as burst or spot-style capacity.

Local utilisation
IDLESpot

Your compute should never sit idle.

Interactive

Introducing FlexGPU : One GPU pool,
many tenants

Add tenant tasks — the control plane schedules the pool

Tenant A8G tasks
2
Tenant B16G tasks
1
Tenant C12G tasks
2
Control planeAdaptive scheduling · GPU resource provisioning · optimal utilisation
GPU pool · 6 × 32G
GPUs2/6
Utilisation88%
16G
12G
GPU 0
12G
8G
8G
GPU 1
Free
GPU 2
Free
GPU 3
Free
GPU 4
Free
GPU 5
Pending queue0
all tasks scheduled
Tenant ATenant BTenant CFreeWasted

Reach

Plug into existing
hyperscaler clusters

No infra change — keep their EKS, AKS & GKE control planes · your GPUs & bare metal join as virtual workers · minimal egress over a private link.

AKS
AKS
Control
Plane
Worker
Worker
VNET
ExpressRoute
GKE
GKE
Control
Plane
Worker
Worker
VPC
Cloud Interconnect
EKS
EKS
Control
Plane
Worker
Worker
VPC
Direct Connect
Private encrypted fabricstack8sstack8sOne private network
Your data centreGPUsBare metalCompute

Positioning

Why stack8s Kubernetes
is different

Most platforms manage clusters — stack8s connects supply and demand

Traditional Kubernetes platforms

Deploy &
operate clusters

stack8s

Connect supply
& demand

Built for infrastructure providers who want to:

Operate a multi-tenant Kubernetes cloud

Expose GPU capacity to external demand

Give customers burst access to other providers

Offer managed AI, data & application services

Reduce idle infrastructure

Build a global compute network — without becoming a hyperscaler

Start monetising

Time to Supercharge
your DataCenter

Stop leaving revenue on the table. Book a call and we'll show you how to turn unused compute into a live income stream — in days, not months.

Backed by
SFC Capital
Rule30
NVIDIA Inception
Member of
CNCF Member
Linux Foundation Member