---
title: "How We Built TLX: A Hardware-Native GPU Compiler for Production ML at Scale"
description: "Triton Low-Level Extensions (TLX) is an extension to Triton that helps engineers optimize GPU performance while preserving Triton’s productive programming model. TLX introduces explicit warp-group and cluster-level orchestration, letting developers adopt new hardware capabilities without sacrificing velocity. Modern GPU architectures increasingly rely on execution-oriented features such as warp specialization, asynchronous memory movement, tensor memory, and cluster-level execution to optimize their performance. TLX has demonstrated performance improvements in RecSys training, RecSys inference, and large-scale LLM pretraining. Across NVIDIA and AMD GPUs, TLX combines state-of-the-art kernels with explicit execution orchestration and model-aware fusion to deliver end-to-end improvements beyond what standalone libraries can achieve."
---

# How We Built TLX: A Hardware-Native GPU Compiler for Production ML at Scale

[Meta](https://yomu.fyi/company/meta) · By Hongtao Yu, Peng Chen, Daohang Shi, Nick Riasanovsky, Lei Wang, Partha Kanuparthy · Nov 10, 2026

**Type:** Announcement

## Summary

Triton Low-Level Extensions (TLX) is an extension to Triton that helps engineers optimize GPU performance while preserving Triton’s productive programming model. TLX introduces explicit warp-group and cluster-level orchestration, letting developers adopt new hardware capabilities without sacrificing velocity. Modern GPU architectures increasingly rely on execution-oriented features such as warp specialization, asynchronous memory movement, tensor memory, and cluster-level execution to optimize their performance. TLX has demonstrated performance improvements in RecSys training, RecSys inference, and large-scale LLM pretraining. Across NVIDIA and AMD GPUs, TLX combines state-of-the-art kernels with explicit execution orchestration and model-aware fusion to deliver end-to-end improvements beyond what standalone libraries can achieve.

## Context

Modern GPU architectures (NVIDIA Hopper, Blackwell, AMD Mi300/Mi350x) introduce execution-oriented features like warp specialization, TMA, TMEM, and cluster-level execution that create a gap between low-level CUDA control and high-level Triton productivity. TLX fills this missing middle by elevating warp groups to first-class execution agents.

## Approach / What changed

TLX extends Triton with explicit warp-group and cluster-level orchestration (MIMW model), enabling developers to coordinate execution agents, asynchronous memory movement, and shared resources on modern GPUs while preserving Triton’s productive programming model.

## Takeaways

- TLX introduces a MIMW (Multiple Instructions, Multiple Warps) programming model that elevates warp groups to first-class execution agents, filling the gap between CUDA’s SIMT and Triton’s SIMB models.
- TLX makes hardware features like WGMMA, TMA, and TMEM accessible through concise annotations, enabling warp-specialized fusion and multi-DSL kernels that would be prohibitively complex in raw CUDA.
- TLX typically delivers performance on-par with hand-tuned CUDA but enables significantly faster iteration, with implementations taking weeks rather than months for complex kernels.

## FAQ

### How does TLX differ from standard Triton?

TLX introduces explicit warp-group and cluster-level orchestration (MIMW model) on top of Triton, letting developers coordinate execution agents, asynchronous memory movement, and shared resources without dropping into low-level CUDA code.

### What performance gains has TLX shown in production?

TLX has demonstrated performance improvements in RecSys training, RecSys inference, and large-scale LLM pretraining at Meta, with kernels that match hand-tuned CUDA performance while enabling faster iteration.

### Which GPU architectures does TLX support?

TLX currently targets NVIDIA Hopper and Blackwell, and AMD Mi300 and Mi350x, with work underway for more architectures including custom accelerators like MTIA.

**Tags:** [GPU](https://yomu.fyi/topic/gpu), [LLMs](https://yomu.fyi/topic/llm), [Machine Learning](https://yomu.fyi/topic/machine-learning), [NVIDIA](https://yomu.fyi/topic/nvidia), [Triton](https://yomu.fyi/topic/triton)

- Source: [Meta](https://engineering.fb.com/2026/11/10/open-source/tlx-a-hardware-native-gpu-compiler-for-production-ml)
- Source URL: https://engineering.fb.com/2026/11/10/open-source/tlx-a-hardware-native-gpu-compiler-for-production-ml
- Ingested by Yomu: 2026-11-10T18:01:08.035Z

[Read original post](https://engineering.fb.com/2026/11/10/open-source/tlx-a-hardware-native-gpu-compiler-for-production-ml)
