From 975ef13560c8ce19c50f3558332d88aff95851fd Mon Sep 17 00:00:00 2001 From: lkcl Date: Wed, 4 May 2022 11:25:57 +0100 Subject: [PATCH] --- openpower/sv/biginteger/analysis.mdwn | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/openpower/sv/biginteger/analysis.mdwn b/openpower/sv/biginteger/analysis.mdwn index c2c2956ec..0ea804a53 100644 --- a/openpower/sv/biginteger/analysis.mdwn +++ b/openpower/sv/biginteger/analysis.mdwn @@ -556,7 +556,11 @@ The very fact that programmers even have to consider multiple implementations and compare their performance is an unavoidable nuisance. SVP64 is supposed to be designed such that only one implementation of any given algorithm is needed. In some ways it is reassuring that -some algorithms just don't fit. +some algorithms just don't fit. Slightly more reassuring is that +Goldschmidt Divide, which uses two multiplications that can be +performed in parallel, would be a much better fit with SVP64 (and +Vector Processing in general), the only downside being that it is +regarded as worthwhile for much larger integers. # Conclusion -- 2.30.2