TrnRT open-source inference engine targets AWS Trainium
What happened
On November 18, 2026, the Hacker News front page covered TrnRT, an open-source LLM inference engine for AWS Trainium. The engine serves an OpenAI-compatible API on a single trn2.48xlarge instance, and its first supported model is GLM-5.3-Flash. The listed model specs: MoE architecture, 320B total parameters, 18B active parameters, FP8, and a 1M-token context. The report covers how the engine is deployed, its API compatibility and its first supported model.
Written by AI from the coverage · updated 3 days ago
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pageShow HN: TrnRT – An open-source LLM inference engine for AWS Trainium
TrnRT 是面向 AWS Trainium 的开源 LLM 推理引擎,可在单台 trn2.48xlarge 上提供 OpenAI 兼容 API,首个支持的模型为 GLM-5.3-Flash(320B MoE、18B 激活、FP8、1M-token 上下文)。
Heat over time
Heat now 3·Comparable peak 9(Nov 18 02:00)·Comparable change over 24 hours -50%
The trend only compares accounts observed without gaps, so its range may be smaller than the current heat. Hover or tap the chart for each hour; the left and right arrow keys step through it.