From: lkcl <lkcl@web>
Date: Sun, 16 Apr 2023 10:32:44 +0000 (+0100)
Subject: (no commit message)
X-Git-Tag: opf_rfc_ls009_v1~29
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=e61935ea7e5a852d2c98668480259c7fc01ab1b2;p=libreriscv.git

---

diff --git a/openpower/sv/remap/appendix.mdwn b/openpower/sv/remap/appendix.mdwn
index 8c67325ab..7e6945e85 100644
--- a/openpower/sv/remap/appendix.mdwn
+++ b/openpower/sv/remap/appendix.mdwn
@@ -98,7 +98,7 @@ If the programmer can find any algorithm
 which has identical triple nesting then the FFT Schedule may be
 used even there.
 
-# 4x4 Matrix to vec4 Multiply Example
+# 4x4 Matrix to vec4 Multiply (4x4 by 1x4)
 
 The following settings will allow a 4x4 matrix (starting at f8), expressed
 as a sequence of 16 numbers first by row then by column, to be multiplied
@@ -140,14 +140,14 @@ equivalent sequence thus is issued:
     fmac f7, f3, f23, f7
 ```
 
-The only other instruction required is to ensure that f4-f7 are
-initialised (usually to zero).
+Hardware should easily pipeline the above FMACs and as long as each FMAC
+completes in 4 cycles or less there should be 100% sustained throughput,
+from the one single Vector FMAC.
 
-It should be clear that a 4x4 by 4x4 Matrix Multiply, being effectively
-the same technique applied to four independent vectors, can be done by
-setting VL=64, using an extra dimension on the SHAPE0 and SHAPE1 SPRs,
-and applying a rotating 1D SHAPE SPR of xdim=16 to f8 in order to get
-it to apply four times to compute the four columns worth of vectors.
+The only other instruction required is to ensure that f4-f7 are
+initialised (usually to zero) however obviously if used as part
+of some other computation, which is frequently the case, then
+clearly the zeroing is not needed.
 
 [[!tag standards]]