# Nanite Performance is Not Better than Overdraw Focused LODs \[TEST RESULTS\]. Epic's Documentation is Dangering Optimization.

**URL:** <https://forums.unrealengine.com/t/nanite-performance-is-not-better-than-overdraw-focused-lods-test-results-epics-documentation-is-dangering-optimization/1263218>\
**Category:** Debugging, Optimization, & Profiling\
**Tags:** unreal-engine, bug-report, Rendering\
**Created:** [August 16, 2023, 7:32pm UTC](https://forums.unrealengine.com/t/nanite-performance-is-not-better-than-overdraw-focused-lods-test-results-epics-documentation-is-dangering-optimization/1263218 "2023-08-16T19:32:57Z")\
**Posts on this page:** 1\
**Showing post:** 2

<div class="post-metadata">

**Author:** ![TheKJ](https://d3kjluh73b9h9o.cloudfront.net/original/4X/2/8/5/28502927f1d97ba1051b3b47ccfe17233bf41ef0.png) [@TheKJ](https://forums.unrealengine.com/u/TheKJ)\
**Post date:** [August 17, 2023, 6:04am UTC](https://forums.unrealengine.com/t/nanite-performance-is-not-better-than-overdraw-focused-lods-test-results-epics-documentation-is-dangering-optimization/1263218/2 "2023-08-17T06:04:15Z")

</div>

BUMP  
(Outdated test, needs overdraw view mode for LODs, we can assume that the LODs did enfact reduce quad overdraw since the results favor LODs )

Here is the ms timing on an even bigger scale test.  
The mesh is 4300-5000 tris(same mesh as before). 10,000+instances.  
With 10 minute made LODs  
Look at the ms render time with LODS.

 ![WithLOD 100000.PNG](https://d3kjluh73b9h9o.cloudfront.net/original/4X/2/a/3/2a35c7be5205290d3811d2a13d0560b3f02e4818.jpeg)

VS Epic’s Nanite, on the **same mesh**.

 ![100000 NANITE.PNG](https://d3kjluh73b9h9o.cloudfront.net/original/4X/f/b/9/fb940a1d6d543e909ac6b4966a96813c388f5be8.jpeg)

With Nanite, it’s rendering 4000-6000 triangles but with LODs it renders 1,000,000+tris and STILL runs faster **without Nanite** (stats via “stat RHI”) . You’ll find a lot more test here showing Nanite causing slower performance.

- This means other, more important features that add to photorealism handle 1,000,000 more triangles better over a full Nanite scene by `1.5ms`.

- That is a test with one mesh? Can you even imagine much better Nanite projects would run be if they used optimized, game focused LODs instead of Nanite with several different kinds of meshes?

- The testing with City Sample and FN performance shows there is no optimizing Nanite unless you have baked environments. And even then the first post shows it doesn’t always render faster. Even without VSMs in the equation.

FIX your documentation before [more studios ruin their games](https://youtu.be/jsggpVOwqQo?t=253).

---

_[View the full topic](https://forums.unrealengine.com/t/nanite-performance-is-not-better-than-overdraw-focused-lods-test-results-epics-documentation-is-dangering-optimization/1263218)._