Trident TCN —
Dilated Causal TCN Language Model
The temporal convolutional variant of Trident on Hailo-10H. 8-layer dilated causal TCN achieving 2,352 FPS. The causal convolution architecture guarantees no future token leakage by construction (unlike the Conv2D attention approximation which requires careful masking). Preferred for production autoregressive generation due to the provable causality guarantee.
Abstract
Background
The temporal convolutional variant of Trident on Hailo-10H. 8-layer dilated causal TCN achieving 2,352 FPS. The causal convolution architecture guarantees no future token leakage by construction (unlike the Conv2D attention approximation which requires careful masking). Preferred for production autoregressive generation due to the provable causality guarantee.
Approach
The model was trained in AxonML (a pure-Rust deep learning framework) and compiled through the Hailo Dataflow Compiler (DFC 5.3.0) targeting Hailo-10H silicon. Post-training INT8 quantization was applied during the DFC compilation pass with production telemetry calibration data. The resulting Hailo Executable Format (HEF) binary executes on Hailo’s fixed-function dataflow architecture with deterministic latency and zero framework overhead at the edge.
Results
On production hardware (Hailo-10H (NexusWatch, FW 5.3.0)), Trident TCN achieves 2,352 FPS with 48.8 °C average die temperature (max 49.0 °C).
Conclusion
Trident TCN is production-ready as a single HEF binary deployed to edge devices with no external dependencies beyond the HailoRT vendor runtime. The model meets real-time latency requirements for its target large language model application.
| Model | Trident TCN |
| Domain | Large Language Model |
| Architecture | Dilated Causal TCN Language Model |
| Target silicon | Hailo-10H |
| Measured on | Hailo-10H (NexusWatch, FW 5.3.0) |
| DFC compiler | 5.3.0 |
| Framework | AxonML v0.6 (pure-Rust, CUDA + CPU backends) |
| Author | Andrew Jewell Sr. · ORCID 0009-0005-2158-7060 |
| Organization | AutomataNexus LLC · Fort Wayne, Indiana |
Executive overview
The temporal convolutional variant of Trident on Hailo-10H. 8-layer dilated causal TCN achieving 2,352 FPS. The causal convolution architecture guarantees no future token leakage by construction (unlike the Conv2D attention approximation which requires careful masking). Preferred for production autoregressive generation due to the provable causality guarantee.
Network I/O
Input: token embeddings [1, 256, seq, 1]. Output: vocabulary logits.
Architecture
Dilated Causal TCN Language Model
8-layer dilated causal TCN, d=256, kernel=3, dilation doubling per layer (1→128). Receptive field: 765 tokens. Residual connections with batch normalization. Vocabulary projection head via 1x1 Conv2D. On Hailo-10H, achieves 2,352 FPS at 48.8 degrees C die temperature — thermally sustainable for continuous generation.
Compilation constraints
All AxonML models targeting Hailo silicon are compiled under the fixed-function dataflow constraints: no dynamic control flow, no variable-length dimensions, all activations representable in INT8 after calibration, and no operations requiring dedicated softmax hardware (replaced with ReLU gating or depthwise convolution equivalents where necessary).
Silicon performance
Measured on production hardware via hailortcli benchmark with 5-second sustained inference. Device: Hailo-10H (NexusWatch, FW 5.3.0).
| Metric | Measured Value |
|---|---|
| FPS (streaming) | 2,351.53 |
| Die Temperature (mean) | 48.75 °C |
| Die Temperature (min) | 47.74 °C |
| Die Temperature (max) | 49.00 °C |
| Quantization | INT8 (post-training, DFC calibration) |
| DFC Compiler | 5.3.0 |
| HailoRT | 5.3.0 |
| Measured On | Hailo-10H (NexusWatch, FW 5.3.0) |
Deployment
Deployed as a single HEF binary. No ONNX runtime, TensorFlow Lite, or Python inference stack required at the edge.
| Target silicon | Hailo-10H |
| Measured on | Hailo-10H (NexusWatch, FW 5.3.0) |
| DFC compiler | 5.3.0 |
| Quantization | INT8 (post-training, production telemetry calibration) |
| Runtime | HailoRT (vendor runtime) |
| Edge platform | Raspberry Pi 5 + Hailo AI HAT+ (M.2 Key M) |
Deployment procedure
Copy the .hef binary to the target device. hailortcli run loads the HEF directly into the Hailo-10H dataflow engine over PCIe. Inference begins immediately with deterministic per-frame latency. No model conversion, graph optimization, or warmup phase required.
References
- Jewell, A. (2026). AxonML: A Pure-Rust Deep Learning Framework for Edge Inference. AutomataNexus LLC. Technical whitepaper.
- Hailo Technologies Ltd. (2024). Hailo Dataflow Compiler User Guide. DFC v5.3.0.
- Hailo Technologies Ltd. (2024). Hailo-10H Product Datasheet.
AutomataNexus LLC · Fort Wayne, Indiana · andrew.jewellsr@automatanexus.com
Andrew Jewell Sr. · ORCID 0009-0005-2158-7060
May 2026 · All rights reserved.