**Notes and Observations**:
-1. TODO
+1. Although costly the alternatives in SIMD-paradigm software result in
+ huge algorithmic complexity and associated power consumption increases.
+ Loop-unrolling compilers are prevalent as is thousands to tens of
+ thousands of instructions.
+2. Core inner kernels of Matrix DCT DFT FFT NTT are dramatically reduced
+ to a handful of instructions.
+3. No REMAP instructions with the exception of Indexed rely on registers
+ for the establishment of REMAP capability.
+4. Future EXT1xx variants and SVP64/VSX will dramatically expand the power
+ and flexibility of REMAP.
+5. The Simple-V Compliancy Subsets make REMAP optional except in the
+ Advanced Levels. Like v3.1 MMA it is **not** necessary for **all**
+ hardware to implement REMAP.
**Changes**
Add the following entries to:
-* the Appendices of Book I
-* Instructions of Book I as a new Section
-* TODO-Form of Book I Section 1.6.1.6 and 1.6.2
+* the Appendices of SV Book
+* Instructions of SV Book as a new Section
+* SVI, SVM, SVM2, SVRM Form of Book I Section 1.6.1.6 and 1.6.2
----------------
# Forms
-Add `SVI, SVM, SVRM` to `XO (26:31)` Field in Book I, 1.6.2
+Add `SVI, SVM, SVM2, SVRM` to `XO (26:31)` Field in Book I, 1.6.2
Add the following to Book I, 1.6.1, SVI-Form
| SVM2 | I | # | 3.0B | svshape2 | REMAP shape instruction (2) |
| SVI | I | # | 3.0B | svindex | REMAP General-purpose Indexing |
+[[!inline pages="openpower/sv/remap/appendix" raw=yes ]]
+
## REMAP pseudocode
Written in python3 the following stand-alone executable source code is the Canonical