# SC26 and Fortran

**URL:** <https://fortran-lang.discourse.group/t/sc26-and-fortran/11140>\
**Category:** Uncategorized\
**Created:** [September 30, 2026, 3:38pm UTC](https://fortran-lang.discourse.group/t/sc26-and-fortran/11140 "2026-09-30T15:38:47Z")\
**Posts on this page:** 1\
**Showing post:** 24

<div class="post-metadata">

**Author:** ![sumseq](https://yyz2.discourse-cdn.com/free1/user_avatar/fortran-lang.discourse.group/sumseq/32/3083_2.png) [@sumseq](https://fortran-lang.discourse.group/u/sumseq)\
**Post date:** [October 7, 2026, 12:02am UTC](https://fortran-lang.discourse.group/t/sc26-and-fortran/11140/24 "2026-10-07T00:02:14Z")

</div>

I feel that the DOE anti-Fortran sentiments was caused by them ordering supercomputers with non-NVIDIA GPUs before the compilers and language were ready for such portability, forcing people to develop Kokkos and Raja - one big example was the (continued) refusal of Intel and AMD to support OpenACC whcih led to a huge obstacle at the time.

I always feel uneasy when I see a Fortran to C++ AI project, or someone spending months converting Fortran to C++ so they can run on GPUs when it’s simply not needed in most cases.

I think having slides showing numpy and Fortran side-by-side with performance on the bottom of each would be great.

---

_[View the full topic](https://fortran-lang.discourse.group/t/sc26-and-fortran/11140)._