Skip to content

Repository files navigation

ExpertMesh

A low-cost, model-aware AMD inference fabric for trillion-parameter sparse MoE models.

CI Release License Status

ExpertMesh asks a different hardware question: instead of buying one enormous HBM supernode or streaming terabytes of weights through CXL, can a sparse model keep each quantized expert permanently resident on inexpensive HBM and move only compact activations?

The current model-targeted path is Kimi K2 Thinking, an official 1T open-weight MoE target. It is deliberately marked targeted, not compatibility-certified until the real checkpoint and target ROCm hardware pass the documented release gates.

Phase 1 explores that design for configurable open-weight sparse MoE models. Model size, expert count, top-k routing, weight format, and active-weight traffic are explicit profile inputs rather than assumptions embedded in the runtime.

ExpertMesh stationary-expert architecture

Why this is different

Conventional serving designs optimize a general-purpose GPU cluster. ExpertMesh deliberately co-designs the runtime and hardware around sparse MoE execution:

  • Stationary experts: quantized expert weights stay on one owning GPU throughout decode.
  • Activation-only transport: RoCE carries routed activations and partial outputs, not model weights.
  • Commodity scale-out: refurbished PCIe HBM cards and used EPYC nodes replace a proprietary 64-GPU scale-up domain.
  • Topology-aware scheduling: work is grouped by expert owner, with local Infinity Fabric preferred over remote transport.
  • Measured hot-expert replication: scarce HBM is spent only after routing traces demonstrate skew.
  • Speculative decoding: a smaller draft model can hide network and kernel latency while the large model verifies batches of tokens.

The defensible engineering work is the data plane: checkpoint ingestion, packed-weight HIP kernels, activation dispatch, expert placement, reduction, and evidence-driven scheduling.

Phase 1 reference system

Target Reference value
Compute nodes 8 refurbished four-GPU EPYC systems
Accelerators 32 × AMD Instinct MI210 64 GB
Aggregate HBM 2 TiB
Aggregate HBM bandwidth 51.2 TB/s theoretical
Inter-node fabric Dual 100/200 GbE RoCE per node
GPU board power 9.6 kW
Estimated acquisition cost $135k–$235k
Active-weight planning range 30–50 GB/token
Bandwidth roofline 1,024–1,707 aggregate token/s

The throughput range is an ideal bandwidth roofline—not a benchmark or performance promise. It excludes communication, routing imbalance, KDA/attention work, KV traffic, dequantization, kernel inefficiency, and synchronization. See the hardware and cost model and source assumptions.

Example sizing profile: Kimi K3

The checked-in Phase 1 profile uses the publicly announced characteristics of Kimi K3 as one demanding example: approximately 2.8T parameters, 896 experts, top-16 routing, and packed four-bit weights. Those values demonstrate the planner at trillion-parameter scale; they do not make the architecture model-specific. Official checkpoint metadata and licensing always override announcement-era assumptions.

What the repository contains

  • A HIP routed-expert correctness kernel for group-scaled signed INT4 weights.
  • A stable C ABI and documented [expert, output, packed_input] layout.
  • A bit-exact NumPy quantizer, dequantizer, routed expert, and sparse MoE reference.
  • Deterministic 896-expert placement across 32 stationary owners.
  • Activation grouping that preserves token and top-k slot identity.
  • Capacity planning and explicitly labeled bandwidth-roofline CLIs.
  • A versioned, checksummed checkpoint format and deterministic expert sharder.
  • A runnable attention → route → dispatch → INT4 SwiGLU → reduce reference runtime.
  • Local and framed-TCP worker transports with numerical parity tests.
  • An OpenAI-compatible chat-completions endpoint and three-container deployment example.
  • A synthetic GPU smoke test plus eighteen CPU unit/integration tests.
  • Architecture, transport, BOM, release gates, sources, and a four-phase roadmap.

Data flow

flowchart LR
    request["Requests"] --> scheduler["Continuous batching and speculation"]
    scheduler --> router["Attention, KDA and top-k router"]
    router --> dispatcher["Group activations by expert owner"]
    dispatcher --> local["Local xGMI hive"]
    dispatcher --> roce["Remote RoCE fabric"]
    local --> workers["Stationary packed-INT4 experts"]
    roce --> workers
    workers --> reducer["Weighted top-k reduction"]
    reducer --> router
Loading

An editable, presentation-ready version is available in FigJam, with repository-native Mermaid and SVG sources.

Quick start without an AMD GPU

python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"

pytest -q
expertmesh-plan configs/phase1-mi210.json
expertmesh-roofline configs/phase1-mi210.json

On Windows, activate with .venv\Scripts\Activate.ps1.

Expected planning output begins with:

32 accelerators; 896 experts; minimum free HBM 17.43 GiB/rank
theoretical aggregate decode roofline: 1024–1707 token/s

Run the end-to-end cluster path

expertmesh-checkpoint .demo --workers 2
expertmesh-serve .demo --workers 2

This creates a tiny deterministic MoE, validates its manifest, assigns experts to two logical workers, runs the full coordinator/worker data path, and exposes POST /v1/chat/completions on port 8000. For separate worker processes or Docker Compose, see the runtime guide. The generated model is a plumbing fixture—not a useful language model.

Build the HIP kernel

The GPU path requires Linux, ROCm, CMake, and a supported AMD accelerator:

cmake -S . -B build -DEXPERTMESH_BUILD_HIP=ON
cmake --build build -j
./build/expertmesh_hip_smoke

The current kernel prioritizes an auditable layout and correctness. It unpacks signed INT4 weights, applies FP16 group scales, executes the routed expert product, and performs a portable workgroup reduction. The post-checkpoint optimization path replaces this with grouped MFMA GEMM and fused gate/up/down projections. See the kernel contract.

Repository map

include/expertmesh/  stable HIP runtime ABI
kernels/hip/        packed INT4 routed-expert kernel and GPU smoke test
src/expertmesh/      checkpoint, runtime, transport, serving, planning and reference kernels
configs/             explicit model and hardware assumption profiles
tests/               synthetic correctness and planning tests
docs/                architecture, transport, BOM, gates, evidence and roadmap

Validation status

Check Status
Python reference tests 18 passing
Ruff static checks Passing
Package build Passing
Planner and roofline CLI Passing
GitHub Actions Passing
HIP source and smoke test Included; requires a ROCm GPU runner
Target checkpoint compatibility Requires an immutable checkpoint and reference runtime
Performance claims Intentionally withheld until measured

Accuracy and scope boundary

This repository does not claim compatibility with any specific trillion-parameter model or that the theoretical roofline will be achieved. Every target—including example targets such as Kimi K3—must provide immutable weights, a compatible license, tensor shapes, quantization semantics, router behavior, attention operators, residual behavior, and bit-exact reference outputs. The mandatory process is documented in release gates.

The architecture is intended for sparse MoE inference. A dense 2.8T model activates nearly all weights per token and has fundamentally different bandwidth economics.

Documentation

Contributing and security

Contributions are welcome, especially reproducible AMD benchmarks, grouped MFMA kernels, ROCm transport experiments, and checkpoint-layout adapters. Read CONTRIBUTING.md before opening a pull request. Please report vulnerabilities through the process in SECURITY.md, not a public issue.

License

Apache-2.0. See LICENSE.

About

Low-cost AMD/ROCm stationary-expert fabric for configurable trillion-parameter sparse MoE inference, with packed-INT4 HIP kernels, activation-only RoCE transport, hardware/BOM design, release gates, and reproducible roofline tooling.

Topics

Resources

Contributing

Security policy

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages