Home

XAIOS Wiki

This directory mirrors selected pages from the live GitHub Wiki. Current source, build/test configuration, docs/MODEL-SUPPORT.json, and docs/PLATFORM-SUPPORT.json take precedence over older Wiki revisions.

XAIOS is an experimental operating system and portable inference-engine foundation. Its current QEMU paths validate deterministic OS/runtime contracts; they do not prove real-model inference or physical-hardware performance.

Delivery Sequence

Only the XAIOS platform workstream is active. Qwen is next, but remains blocked until XAIOS reaches its completion gate. Later model workstreams are not active unless the maintainer explicitly reprioritizes them.

Order	Workstream	Project status	Entry gate
1	XAIOS	In Progress	Finish the core OS, portable engine, model-v2 integration, platform services, hardware readiness, and release gates.
2	Qwen 3.6 27B Support	Blocked	Starts only after the XAIOS completion gate.
Later	Kimi K3 Support	Backlog	Queued behind XAIOS and Qwen unless explicitly reprioritized.
Later	DeepSeek V4 Flash 0731 Support	Blocked	Also blocked on authoritative release and source verification.
Later	GLM 5.2 Support	Backlog	Queued behind XAIOS and Qwen unless explicitly reprioritized.

Current Boundaries

• The model-v1 QEMU path is a deterministic fixture, not a transformer.
• xaios.model.v2 and portable adapter/backend APIs are interface foundations.
• No listed real model has passed tokenizer, logits, deterministic decode, and physical-hardware acceptance gates.
• QEMU is correctness and ABI evidence only.
• Performance claims require immutable artifacts under the benchmark contract.
• xaiosctl Phase 2 is QEMU/OpenSSH fixture-tested with role-mapped keys, revocation, config transactions, host-key rotation, redacted audit and typed storage lifecycle administration.
• Signed ModelFS supports dynamic registration, resumable SFTP, cleanup/reuse, verification, atomic activation, scrub/quarantine and free-only trim under QEMU. Concurrent macOS/Debian clients pass against one guest. QEMU VirtIO uses interrupt-driven block/network completions, event-index suppression, indirect descriptors and eight-request block batching. A focused emulated-NVMe gate verifies identify/write/flush/read and host backing bytes; production multiqueue and physical storage validation remain open.
• Runtime-sized NUMA/CPU/cpuset state and CPU-assigned worker threads pass QEMU; a focused TCG gate validates SMP and NUMA metadata with 130 emulated CPUs, while hosted cpuset tests cover 4,097 CPU IDs. EL0 create/join/cancel/exit, asynchronous DNS, IPv4/IPv6 reassembly, and SACK-aware TCP pass QEMU gates.
• x86_64 starts MADT-discovered application processors, dispatches IPI work, validates controlled exception and local-APIC timer interrupts, performs a ring-3 syscall round trip, validates runtime-sized XSAVE and ACPI parsing, and operates modern VirtIO block DMA/MSI-X plus network TX. Full ARM-service parity on x86 remains open: complete userspace/thread services, receive networking/SSH, mounted filesystems, x86 NVMe, security, AI Cell and telemetry are not yet integrated.
• Model loading, cluster and inference-service administration remains gated.

Start Here

• Model Support Roadmap
• Platform Support
• Qwen CPU Inference Status
• SSH Status
• Repository README
• Implementation roadmap
• Distributed server plan
• xaiosctl reference
• Large-model upload status
• Benchmark contract
• GitHub Project