From 83af7ed1bc6fc4e6a8c362761888dd5d357ce788 Mon Sep 17 00:00:00 2001
From: lkcl <lkcl@web>
Date: Sun, 24 Oct 2021 12:31:21 +0100
Subject: [PATCH]

---
 3d_gpu/architecture/dynamic_simd/slice.mdwn | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/3d_gpu/architecture/dynamic_simd/slice.mdwn b/3d_gpu/architecture/dynamic_simd/slice.mdwn
index aa4890cce..7dbd00b63 100644
--- a/3d_gpu/architecture/dynamic_simd/slice.mdwn
+++ b/3d_gpu/architecture/dynamic_simd/slice.mdwn
@@ -258,4 +258,7 @@ Combining a slice and b at each elwidth gives:
     0b10  B7B6Ae AdAc|B5B4AaA9 A8|B3B2A6 A5A4|B1B0A2A1A0
 
 This complex-looking result is quite straightforward: SimdShape
-is capable of expressing it.  
+is capable of expressing it.  The issue is, however: how
+to convert (adapt) the two inputs so that they fit this resultant
+PartitionPoint layout, given that neither input was in this
+format?
-- 
2.30.2