Skip to content
André Borchert edited this page Jun 13, 2026 · 3 revisions

FAQ

Purpose

This page answers common questions about OSAI.

Is OSAI a Linux distribution?

No. OSAI is not a Linux distribution.

Is OSAI a BSD fork?

No. OSAI is not a BSD fork.

Does OSAI use GPUs?

No GPU is required. OSAI targets CPU-only AI and has no CUDA, Metal, or GPU acceleration dependency.

Why CPU-only AI?

CPU-only AI is deployable on ordinary servers and appliances, avoids vendor GPU runtime dependency, and lets the OS control cores, memory placement, networking, and build/test work in one resource model.

Can OSAI really beat Linux/BSD?

Only measured results can answer that. OSAI has design targets for a narrow workload: CPU-only embedded app agents with explicit core ownership, memory arenas, and low-latency networking. It should not claim broad superiority over Linux/BSD.

What does near-zero migration mean?

It means hot AI workers should remain on their leased cores unless the service explicitly changes placement.

Why start with QEMU on macOS?

QEMU on macOS gives a reproducible correctness target for boot, memory, devices, and kernel basics. It is not a performance target.

Why Intel Desktop before Xeon?

Intel Desktop is accessible and exposes real P-core/E-core, NVMe, NIC, and sustained-performance behavior before the complexity of NUMA servers.

Why ARM/NVIDIA last?

ARM/NVIDIA N1X-compatible work needs the OS architecture to be stable before AArch64 boot, GIC, SMMU, and cluster policies are added.

What language will OSAI be written in?

Early kernel code should be freestanding C99 plus minimal architecture assembly. C++ is allowed in userspace, runtime, and AI components.

Can apps still use C/C++?

Yes. OSAI should support static C/C++ services early.

Can apps use C/C++?

Yes. This is the same policy as above: C and C++ are expected for early userspace, runtime, and application services.

Will Swift be supported later?

Possibly, but Swift is not an early kernel or bring-up requirement.

How will Git/code-changing agents be kept safe?

Through explicit permissions, scoped workspaces, build/test sandboxes, diff review, signed updates, audit logs, and rollback.

Related Pages

Clone this wiki locally