From 0bb9d57fcd78ae9bc0fd27d75767f24f3743bff4 Mon Sep 17 00:00:00 2001 From: lkcl Date: Mon, 5 Sep 2022 13:27:18 +0100 Subject: [PATCH] --- openpower/sv/remap.mdwn | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/openpower/sv/remap.mdwn b/openpower/sv/remap.mdwn index 6232c2230..04f5546bf 100644 --- a/openpower/sv/remap.mdwn +++ b/openpower/sv/remap.mdwn @@ -48,7 +48,7 @@ latency, so should realistically be used only where it is worthwhile. Commonly-used patterns such as Matrix Multiply, DCT and FFT have helper instruction options which make REMAP easier to use. -There are three types of REMAP: +There are four types of REMAP: * **Matrix**, also known as 2D and 3D reshaping, can perform in-place Matrix transpose and rotate. @@ -56,12 +56,20 @@ There are three types of REMAP: Power-2 RADIX * **Indexing**, for any general-purpose reordering, also includes limited 2D reshaping. +* **Parallel Reduction**, for scheduling a sequence of operations + in a Deterministic fashion, in a way that may be parallelised, + to reduce a Vector down to a single value. Best implemented on top of a Multi-Issue Out-of-Order Micro-architecture, REMAP Schedules are 100% Deterministic **including Indexing** and are designed to be incorporated in between the Decode and Issue phases, directly into Register Hazard Management. +Parallel Reduction is unusual in that it requires a full vector array +of results (not a scalar) and uses the rest of the result Vector for +the purposes of storing intermediary results. *As these intermediary +results are Deterministically computed they may be useful*. + # Basic principle * normal vector element read/write of operands would be sequential -- 2.30.2