# Understanding Tullio performances

**URL:** <https://discourse.julialang.org/t/understanding-tullio-performances/129760>\
**Category:** Performance\
**Tags:** tullio, jax\
**Created:** [June 9, 2025, 10:21am UTC](https://discourse.julialang.org/t/understanding-tullio-performances/129760 "2025-06-09T10:21:00Z")\
**Posts on this page:** 1\
**Showing post:** 13

<div class="post-metadata">

**Author:** ![yolhan\_mannes](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/yolhan_mannes/32/220485_2.png) [@yolhan\_mannes](https://discourse.julialang.org/u/yolhan_mannes)\
**Post date:** [June 10, 2025, 4:41am UTC](https://discourse.julialang.org/t/understanding-tullio-performances/129760/13 "2025-06-10T04:41:23Z")

</div>

Ah yeah the size of the array may be too low and simd and threads are not composing well.

However we already beat JAX with both (KA or LV)🥳

---

_[View the full topic](https://discourse.julialang.org/t/understanding-tullio-performances/129760)._