Hacker GitHub

← All repos

NVIDIA/TensorRT

View on GitHub ↗ ★ 13,443 C++

NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.

4 mentions on Hacker News by 4 people: 1 story, 3 comments · first 2019-07-02 · latest 2023-12-18

Mentions per year 4 mentions · busiest year 2023 (3)
20082026
AllStoriesComments NewestTopOldest
  1. …recognize that momentum is an advantage of being first mover. And it isn't one to scoff at. [0] https://github.com/NVIDIA/TensorRT
    comment by godelski · 2023-12-18 · in AMD MI300X 30% higher performance than Nvidia H100, even with optimized stack
  2. …many of the larger "production" inference tools use compilers: - https://github.com/openai/triton - https://github.com/NVIDIA/TensorRT TVM and other compiler-based approaches seem to really perform really well and make supporting different…
    comment by jmorgan · 2023-09-26 · in Ollama for Linux – Run LLMs on Linux with GPU Acceleration
  3. https://github.com/NVIDIA/TensorRT/issues/982 Looks like tensorRT does work with 30x0, but I couldn't find much.
    comment by nkingsy · 2023-09-08 · in NVIDIA introduces TensorRT-LLM for accelerating LLM inference on H100/A100 GPUs