1735671969947803827

2026-08-25 09:37:49 -04:00
1 parent 8103e2c8de
commit bca5513b5c
@@ -0,0 +1,39 @@
+---
+title: "@NOTimothyLottes They design one API for each separate use case instead of making general purpose flexible SIMD hardware and API for the developers to implement what ever algorithm they need."
+type: archive
+source: twitter
+source_url: "https://x.com/SebAaltonen/status/1735671969947803827"
+author: "Sebastian Aaltonen"
+handle: SebAaltonen
+post_id: "1735671969947803827"
+date: 2023-12-15
+archived: 2026-08-25
+draft: false
+tags:
+  - archive
+  - twitter
+  - NOTimothyLottes
+description: "@NOTimothyLottes They design one API for each separate use case instead of making general purpose flexible SIMD hardware and API for the developers to implement what ever algorithm they need."
+in_reply_to: ""
+parent_post_id: "1735631826427732305"
+---
+
+## Source
+
+- URL: https://x.com/SebAaltonen/status/1735671969947803827
+- Author: Sebastian Aaltonen (@SebAaltonen)
+- Posted: 2023-12-15 14:43:39
+
+## Branch
+
+**1/**
+
+@NOTimothyLottes They design one API for each separate use case instead of making general purpose flexible SIMD hardware and API for the developers to implement what ever algorithm they need. CPU side got this right. PS2 had correct design, but Nvidia choose hardware T&L for desktop :(
+
+**2/**
+
+@NOTimothyLottes I kind of understand hardware T&L back then because it was a new field and programmable hardware didn't exist. But now we have massive amount of wide SIMDs and bandwidth and big caches too. We can do better!
+
+## Related
+
+- Spine: [[archive/threads/NOTimothyLottes/2023-12-15-long-thread]]
@@ -7,8 +7,8 @@ author: "NOTimothyLottes"
 handle: NOTimothyLottes
 post_id: "1735622924571201674"
 date: 2023-12-15
-archived: 2026-08-23
-draft: true
+archived: 2026-08-25
+draft: false
 tags:
  - archive
  - twitter
@@ -123,6 +123,8 @@ and make sure to bring max values out of denormal before using rcp.

 [22] Love HW people, but they opt for a $ train even as it is derailing: they need to build HW, make it faster, it's what they do. But leave a trail of back-compat debris TS/ROV/GS/etc after each gfx fad. Vote with your engineering, general purpose CS, just say no to black boxes.

+Branches: [[archive/threads/NOTimothyLottes/2023-12-15-long-thread/2023-12-15-SebAaltonen-they-design-one-api-for-each-separate-use-case]]
+
 **25/**

 [23] Scaling TAAs want sparse data each frame, more like a low-frequency pixel control cage displacing a high-frequency reprojected feedback. No pixel centers any more, no need to interpolate before shading. Rather want striped data, so the cachelines group spaced-out samples ...
@@ -1,32 +0,0 @@
---
-title: "@bmcnett That line of thinking is universal solution to most self-inflicted problems of the modern era."
-type: archive
-source: twitter
-source_url: "https://x.com/NOTimothyLottes/status/1857820858162753661"
-author: "NOTimothyLottes"
-handle: NOTimothyLottes
-post_id: "1857820858162753661"
-date: 2024-11-16
-archived: 2026-08-23
-draft: true
-tags:
-  - archive
-  - twitter
-  - NOTimothyLottes
-description: "@bmcnett That line of thinking is universal solution to most self-inflicted problems of the modern era."
-in_reply_to: "1857819880600531261"
---
-
-## Source
-
- URL: https://x.com/NOTimothyLottes/status/1857820858162753661
- Author: NOTimothyLottes (@NOTimothyLottes)
- Posted: 2024-11-16 16:19:43
-
-> @NOTimothyLottes Variable line width ASCII text files made a lot of sense when storage was 100KB or 100MB or 100GB, but now that it's in the terabytes, why even bother
-
-## Thread
-
-**1/**
-
-@bmcnett That line of thinking is universal solution to most self-inflicted problems of the modern era.