01 The short version
For organisations that need a non-NVIDIA path, or whose mainland customers require a domestic stack end to end. Expect to spend time on MindSpore/CANN compatibility rather than on the defaults that work elsewhere.
- Ascend inference works, but assume a porting effort measured in weeks and benchmark before committing.
- Strong presence in markets where the other hyperscalers are thin — parts of Africa and the Middle East.
- Procurement and compliance posture is the usual reason clients end up here, not raw price.
02 What we do here
BUILD
Inference on Huawei
Open-weight models running on Huawei capacity, gateway routing in front, full cost and usage records underneath, the networking and IAM to hold it together. Deployed in your account with Terraform you keep.
RUN
Operate it monthly
Capacity planning against Huawei quota and pricing, model rotations, incident response, and a monthly report that reconciles against your actual Huawei invoice.
FOUNDATION
Accounts, network, IAM
If the Huawei footprint does not exist yet, or exists as one account somebody made in 2021, we build the landing zone underneath before anything else goes on top.
03 The others
We are not loyal to any of them. The right platform is the one where your data already is, your latency is acceptable and the GPUs are actually available this quarter — and that answer changes.