Services & engineering
Engineers in Kathmandu, on your project.
Help from the people who run the GPUs: sizing capacity, moving work from another cloud, getting a model trained and into production, and setting up GPUs for a class or a lab.
Coming soonEnterprise early access is open
How it works
Three steps.
- 01
Tell us the goal
A migration, a first model in production, a GPU lab for a class.
- 02
Agree the scope
What we'll do, by when, quoted in rupees.
- 03
Work with us
Engineers in Kathmandu, on Nepal time, alongside your team.
What you get
Services & engineering, in detail.
Capacity planning
Tell us the model, the data and the deadline. We size the GPUs, storage and term, and quote it in rupees.
Migration from another cloud
We help move training, data and serving to the Kathmandu Valley and check it runs the same here.
From training to production
Hands-on help to get a fine-tune running and the model behind an endpoint.
GPUs for teaching and research
GPU access for a class, a lab or a department, with an isolated slice per person.
On Nepal time
Engineers in Kathmandu, working on Nepal Standard Time (UTC+05:45).
Under the hood
The key numbers.
- Nepal Standard Time
- UTC+05:45
- Reply time
- Within one working day
- Nepal Standard Time
- UTC+05:45
- Reply time
- Within one working day
- GPU
- NVIDIA H200 NVL
- GPU memory
- 141 GB
- Memory bandwidth
- 4.8 TB/s
- Currency
- Nepalese rupees (NPR)
- Ways to pay
- eSewa, Khalti, bank transfer or invoice
Billing
Scoped and quoted in NPR
Commitment: per engagement. Rates are being finalised; every request gets a firm quote in rupees.
How billing worksQuestions
About services & engineering.
Do we need services to use the GPUs?
No. They are there for teams who want help; many will not need them.
How is it priced?
Each engagement is scoped and quoted in rupees before it starts.
Start with services & engineering.
Tell us what you want to run. It takes about a minute.