XERJI / Inference platform

Your models.
Your ambition.
Your compute.

An inference network in development, with a platform to manage workloads, access GPU capacity and offer compute across connected locations.

Choose a platform
Illustrative view of GPU compute infrastructure
01 / Platform access

Two routes.
One XERJI relationship.

GPU1 and GPU2 are planned access points to the XERJI compute offering. Share your workload requirements or available capacity while platform access is being prepared.

Primary platformIn preparation

GPU1

A planned entry point for managing AI workloads and accessing GPU compute through the XERJI platform.

  • Inference and workload management
  • GPU capacity access
  • Deployment planning
Register your interest
Additional platformIn preparation

GPU2

An additional platform route, designed to broaden the compute options available to customers and connect more capacity to workload demand.

  • Additional capacity options
  • Platform choice
  • Workload-led selection
Register your interest

Workloads and capacity.
In one network.

We’re developing a platform for teams to manage workloads and for providers to offer compute, connected to the future XERJI inference network.

01

Manage workloads.

Plan deployments around models, memory, location and usage. Bring resource requirements and scaling needs into one workload view.

02

Access compute.

Match inference and AI workloads with suitable GPU capacity, deployment regions and commercial terms.

03

Offer compute.

Bring available GPU capacity to the network. Start with your hardware, location, connectivity and availability.

Discuss your workloadOffer compute

A few practical questions

Before you
choose.

Can I deploy GPUs immediately?

The XERJI platform routes are being prepared. Register your interest in GPU1 or GPU2 to discuss your workload or available compute. Registration does not reserve capacity or require payment.

What is the difference between GPU1 and GPU2?

They are separate entry points for XERJI’s planned compute platforms. The final hardware, regions, pricing and service terms will be shown by each platform when it becomes available.

Can we discuss dedicated or future capacity?

Yes. Share your GPU quantity, preferred region, timing and duration with our team. Any offer will depend on confirmed capacity and agreed commercial terms.

How can I offer compute?

Contact us with your GPU type and quantity, location, networking, availability and preferred terms. We’ll discuss suitability and the path to provider onboarding as the network develops.

Do I need to know which GPU I want?

No. Start with your model, memory needs, workload and usage expectations. We can use those requirements to frame the capacity discussion.