RegNetX-400MF (ONNX) – Renesas X5H

Introduction

This repository hosts RegNetX-400MF, targeting the Renesas R-Car X5H platform for image-classification inference on the NPX6 NPU.

  • Model Architecture: RegNetX-400MF β€” a member of the RegNet design-space family (Radosavovic et al.), using a simple regular network structure with a linear per-block width parameterization, at a ~400M-FLOP compute budget
  • Source Model: open-mmlab/mmpretrain
  • Task: image-classification (dataset: imagenet-1k)

Deployment Flow

The FP32 ONNX model is auto-cast to INT8 by the Renesas MWMX toolchain at compile time β€” no separate quantization step is required.

regnetx_400mf_8xb128_in1k.onnx (FP32)
        β”‚
        └─▢  MWMX Runtime  ──▢  INT8 auto-cast  ──▢  NPX6 NPU

Provided Artifacts

Artifact Status Notes
FP32 (ONNX) ⏳ Pending fp32/regnetx_400mf_8xb128_in1k.onnx β€” to be added; will be auto-cast to INT8 by the MWMX toolchain at compile time (see Deployment Flow above); no separate INT8 file will be shipped

Performance

Measured on Renesas R-Car X5H via the MWMX runtime (APM80 ship-performance CI pipeline).

Benchmark configuration: Single NPU Β· Batch size: 1 Β· Input: 3 Γ— 224 Γ— 224

AI Cores Runtime Precision Device Latency (ms) Type
1 MWMX Runtime INT8 (auto) X5H Β· 1Γ— NPU Β· 1 Core Β· 850 MHz 1.31 Measured
3 MWMX Runtime INT8 (auto) X5H Β· 1Γ— NPU Β· 3 Core Β· 850 MHz 1.04 Measured
4 MWMX Runtime INT8 (auto) X5H Β· 1Γ— NPU Β· 4 Core Β· 850 MHz 1.12 Measured
6 MWMX Runtime INT8 (auto) X5H Β· 1Γ— NPU Β· 6 Core Β· 850 MHz 1.16 Measured
12 MWMX Runtime INT8 (auto) X5H Β· 1Γ— NPU Β· 12 Core Β· 850 MHz 1.57 Measured

Accuracy

TBD β€” not yet measured/published for this repo.

ONNX Runtime – Renesas EP v2.1.0 (MWMX 2.2 ENG4)

Measured on R-Car X5H with ONNX Runtime + Renesas Execution Provider (INT8 auto-cast); nodes the NPU cannot run fall back to the CPU EP. Batch size 1, 1 NPU. Source: ORT Renesas EP test report, status 2026-09-30. This section supersedes earlier ORT figures in this card.

Component AI Cores Latency (ms) Throughput (fps) NPU Inference (ms) NPU % CPU + Overhead % Portable pkg NPU-only (ms)
Full model 1 1.985 503.8 1.308 65.9 % 34.1 % 1.28818
Full model 12 2.863 349.3 1.62 56.6 % 43.4 % 1.57068

12-core latency is not always lower than 1-core: small models are dominated by CPU-side overhead.

Graph partitioning (NPU vs CPU nodes)

Component Total Nodes NPU Nodes CPU Nodes (Q/DQ inserted) NPU % CPU %
Full model 165 164 1 (1) 99.4 % 0.6 %


Runtime Details

MWMX Runtime

  • Engine: Renesas MWMX (Middleware MX) native inference runtime
  • Input format: FP32 ONNX (compiled by the MWMX toolchain)
  • NPU execution precision: INT8 (auto-cast by MWMX toolchain)
  • Execution target: NPX6-48K NPU on R-Car X5H

Prerequisites

To run inference on Renesas R-Car X5H, you need:

  1. Renesas R-Car X5H board with NPX6 NPU
  2. Renesas MWMX Runtime
  3. Hugging Face CLI to download the model

Download

hf download Renesas/RegNet-400MF-ONNX --repo-type=model --include "fp32/*"

Benchmark Methodology

  • HIL runs: Hardware-in-the-loop β€” measured on physical R-Car X5H silicon via the MWMX runtime (metawaremx_runtime CI pipeline, "APM80" ship-performance target)
  • Precision: FP32 ONNX input; INT8 execution (auto-cast by MWMX)
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support