# How CUDA Won by NOT being a Standard | SCALE

**URL:** https://fortran-lang.discourse.group/t/how-cuda-won-by-not-being-a-standard-scale/10911
**Category:** Uncategorized
**Created:** [May 17, 2026, 9:24pm UTC](https://fortran-lang.discourse.group/t/how-cuda-won-by-not-being-a-standard-scale/10911 "2026-05-17T21:24:05Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![ivanpribec](https://yyz2.discourse-cdn.com/free1/user_avatar/fortran-lang.discourse.group/ivanpribec/32/3290_2.png) [@ivanpribec](https://fortran-lang.discourse.group/u/ivanpribec)
#### Post date: [May 17, 2026, 9:24pm UTC](https://fortran-lang.discourse.group/t/how-cuda-won-by-not-being-a-standard-scale/10911/1 "2026-05-17T21:24:05Z")

</div>

> **[How CUDA Won by NOT being a Standard | SCALE - Accelerate your CUDA Development](https://scale-lang.com/posts/2026-05-13-how-cuda-won-by-not-being-a-standard?utm_source=linkedin&utm_medium=organic_social&utm_campaign=michael_sondergaard&utm_term=generic&utm_content=text_only)**
>
> SCALE (re)compiles CUDA code for a broad range of GPUs to increase performance.

---

<div class="post-metadata">

### Author: ![ivanpribec](https://yyz2.discourse-cdn.com/free1/user_avatar/fortran-lang.discourse.group/ivanpribec/32/3290_2.png) [@ivanpribec](https://fortran-lang.discourse.group/u/ivanpribec)
#### Post date: [May 27, 2026, 7:28pm UTC](https://fortran-lang.discourse.group/t/how-cuda-won-by-not-being-a-standard-scale/10911/2 "2026-05-27T19:28:29Z")

</div>

The SCALE team published two follow-up articles:

- [The brain still needs the hammer: Why compilers matter MORE in the agent era, not less | SCALE - Accelerate your CUDA Development](https://scale-lang.com/posts/2026-05-19-the-brain-still-needs-the-hammer)
- [CUDA was always cross-platform | SCALE - Accelerate your CUDA Development](https://scale-lang.com/posts/2026-05-27-cuda-was-always-cross-platform)

The second one finishes with the thought,

> … whichever hardware wins, the people building on top of it will want their existing code to keep working, their agents to keep generating code they can trust, and their substrate to stay stable across the change. […]

This reminds me of that Alan Perlis epigram,

> Adapting old programs to fit new machines usually means adapting new machines to behave like old ones.

To me it makes a lot of sense to have CUDA compile on other hardware.

I’m not sure which company was the second to produce a FORTRAN compiler after IBM, I imagine it would have been a company like DEC, Univac, CDC, Burroughs, …?

---

<div class="post-metadata">

### Author: ![Beliavsky](https://avatars.discourse-cdn.com/v4/letter/b/ba8739/32.png) [@Beliavsky](https://fortran-lang.discourse.group/u/Beliavsky)
#### Post date: [May 27, 2026, 8:21pm UTC](https://fortran-lang.discourse.group/t/how-cuda-won-by-not-being-a-standard-scale/10911/3 "2026-05-27T20:21:47Z")

</div>

ChatGPT [said](https://chatgpt.com/share/6a17524c-37ac-832f-8b6c-142c3f64eea9) CDC was 2nd after IBM to create a Fortran compiler, referencing the BCS [List of FORTRAN Implementations 1957 - 1967](https://fortran.bcs.org/2007/jubilee/implementations.php).
