TRT-LLM
Last updated
, generated by
Sumble
Explore more → **TRT-LLM**
What is TRT-LLM?
TRT-LLM (TensorRT-LLM) is an open-source library developed by NVIDIA that provides developers with tools to define, optimize, and execute large language models (LLMs) with high performance on NVIDIA GPUs.
What other technologies are related to TRT-LLM?
TRT-LLM Competitor Technologies
TRT-LLM Complementary Technologies
NVIDIA Triton
NVIDIA Triton Inference Server frequently integrates TRT-LLM as a backend engine to serve models in production, making them strongly complementary.
TensorRT
TensorRT provides the underlying deep learning inference optimizer and runtime primitives upon which TRT-LLM is built.
CUDA
CUDA is the foundational parallel computing platform and programming model by NVIDIA required to execute TRT-LLM on GPUs.
Number of organizations that mention technology
ⓘ Tap on a tech to explore matching organizations
Which job functions commonly mention TRT-LLM?
Machine Learning
45 Machine Learning jobs mention TRT-LLM
View 45
jobs on Sumble
Performance Engineer
6 Performance Engineer jobs mention TRT-LLM
View 6
jobs on Sumble
Solutions Architect
35 Solutions Architect jobs mention TRT-LLM
View 35
jobs on Sumble
AI Engineer
33 AI Engineer jobs mention TRT-LLM
View 33
jobs on Sumble
Software Development Engineer in Test
28 Software Development Engineer in Test jobs mention TRT-LLM
View 28
jobs on Sumble
Which organizations are mentioning TRT-LLM?
Summary powered by 
Sumble
Sumble Find the right accounts, contact, message, and time to sell
Whether you're looking to get your foot in the door, find the right person to talk to, or close the deal — accurate, detailed, trustworthy, and timely information about the organization you're selling to is invaluable.
Use Sumble to: