# Serving ML Model Pipelines on NVIDIA Triton Inference Server with Ensemble Models

**URL:** <https://forums.developer.nvidia.com/t/serving-ml-model-pipelines-on-nvidia-triton-inference-server-with-ensemble-models/245913>\
**Category:** Technical Blog\
**Created:** [March 13, 2023, 2:00pm UTC](https://forums.developer.nvidia.com/t/serving-ml-model-pipelines-on-nvidia-triton-inference-server-with-ensemble-models/245913 "2023-03-13T14:00:24Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![jwitsoe](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/jwitsoe/32/16591_2.png) [@jwitsoe](https://forums.developer.nvidia.com/u/jwitsoe)\
**Post date:** [March 13, 2023, 2:00pm UTC](https://forums.developer.nvidia.com/t/serving-ml-model-pipelines-on-nvidia-triton-inference-server-with-ensemble-models/245913/1 "2023-03-13T14:00:24Z")

</div>

Originally published at: [https://developer.nvidia.com/blog/serving-ml-model-pipelines-on-nvidia-triton-inference-server-with-ensemble-models/](https://developer.nvidia.com/blog/serving-ml-model-pipelines-on-nvidia-triton-inference-server-with-ensemble-models/)  
   
Learn the steps to create an end-to-end inference pipeline with multiple models using NVIDIA Triton Inference Server and different framework backends.

---

<div class="post-metadata">

**Author:** ![flexwang](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/flexwang/32/266174_2.png) [@flexwang](https://forums.developer.nvidia.com/u/flexwang)\
**Post date:** [July 13, 2023, 10:32pm UTC](https://forums.developer.nvidia.com/t/serving-ml-model-pipelines-on-nvidia-triton-inference-server-with-ensemble-models/245913/2 "2023-07-13T22:32:06Z")

</div>

Thanks for the detailed tutorial, very useful!

However, this doesn’t seem like an apple-to-apple comparison. What if we do the pre and post processing locally using GPU, then the latency should be the same?