Services
On Premise Deployment
Deploy the complete SambaNova inference stack as hardware you own, on your premises or in a colocation facility we help you arrange. Full control, with our ongoing support.
90 days
Deployment timeline
1 rack
Start small, scale up
Air-cooled
No liquid cooling required
The complete inference stack, owned by you.
For organisations that need full physical control of their AI infrastructure, we deploy SambaNova hardware as a stack you own outright. Air-cooled, with no liquid cooling required. Start with a single rack and scale as needed, with dedicated support throughout.
- Complete inference stack. Hardware and software, ready to run.
- Air-cooled, single rack. No liquid cooling required, start with one rack.
- Unlimited models. Deploy anything SambaNova supports.
- 90-day deployment. Faster than GPU-based alternatives.
- Dedicated support. From installation through ongoing operation.
Your hardware, your choice of location.
On your premises
Deploy the full stack in your own data centre. You own the hardware and software outright, with complete physical control. We handle installation, configuration, and ongoing support and maintenance.
In a colocation facility
If you want ownership without housing the hardware yourself, we help you arrange colocation in a suitable facility and support it the same way. You own the stack, without the data-centre overhead.
When ownership makes sense.
Owning your infrastructure suits organisations with sustained, high-volume workloads, strict data governance requirements, or a need for complete physical control over where and how AI runs. For regulated sectors in particular, owning the stack removes the last remaining dependency. The infrastructure itself is yours.
Interested in on premise deployment?
Let's discuss hardware requirements, deployment timeline, and ongoing support for your installation.
Discuss your deployment