minor correction

[crowdsupply.git] / updates / 005_2018dec14_simd_without_simd.mdwn

diff --git a/updates/005_2018dec14_simd_without_simd.mdwn b/updates/005_2018dec14_simd_without_simd.mdwn
index fb75a04f0ff941b1e803a17a3a165ed82ec7addc..c6e456c774b7254d90955758fa3f66f41a02c051 100644 (file)
--- a/updates/005_2018dec14_simd_without_simd.mdwn
+++ b/updates/005_2018dec14_simd_without_simd.mdwn
@@ -141,4 +141,10 @@ from *both* FUs.
 The primary focus is on 32-bit (single-precision floating-point) performance
 anyway, for 3D, so if 64-bit operations happen to have half the number of
 Reservation Stations / Function Units, and block more often, we actually
-don't mind so much.  
+don't mind so much.  Also, we can still apply the same "banks" trick on
+the Register File, except this time with 4-way multiplexing on 32-bit
+wide banks, and 4x4 crossbars on the bytes:
+
+{{register_file_multiplexing.jpg}}
+
+