From ada053810057b5b549cb888d4f746bc831433cab Mon Sep 17 00:00:00 2001 From: lkcl Date: Sun, 24 Jul 2022 12:07:38 +0100 Subject: [PATCH] --- openpower/sv/vector_isa_comparison.mdwn | 23 ++++++++++------------- 1 file changed, 10 insertions(+), 13 deletions(-) diff --git a/openpower/sv/vector_isa_comparison.mdwn b/openpower/sv/vector_isa_comparison.mdwn index 10a6a7a0f..cbcfe439f 100644 --- a/openpower/sv/vector_isa_comparison.mdwn +++ b/openpower/sv/vector_isa_comparison.mdwn @@ -96,19 +96,6 @@ It is clear that it is expected to deploy Multi-Issue to achieve high performance, which is a much cleaner approach that has not resulted in ISA poisoning such as that suffered by x86 (AVX). -# Under investigation - -* [ETA-10](http://50.204.185.175/collections/catalog/102641713) - an extremely rare Scalable Vector Architecture from 1986, - similar to the CDC Cyber 205. - Only 25 machines were ever delivered. Page 3-220 of its ISA - shows that it had Predicate Masks and Horizontal Reduction. - Appendix H-1 shows it is likely a Memory-to-Memory Vector - Architecture, and overcame the penalties normally associated - with this by adding an explicit "Vector operand forwarding/chaining" - instruction (Page 3-69). It is however clearly Scalable, up to Vector - elements of 2^16. - # Actual 3D GPU Architectures and ISAs (all SIMD) All of these are not Vector ISAs, they are SIMD ISAs. @@ -142,6 +129,16 @@ All of these are not Vector ISAs, they are SIMD ISAs. Mitch on direct contact with him. It is a different approach from the others, which may be termed "Cray-Style Horizontal-First" Vectorisation. 66000 is a *Vertical-First* Vector ISA. +* [ETA-10](http://50.204.185.175/collections/catalog/102641713) + an extremely rare Scalable Vector Architecture from 1986, + similar to the CDC Cyber 205. + Only 25 machines were ever delivered. Page 3-220 of its ISA + shows that it had Predicate Masks and Horizontal Reduction. + Appendix H-1 shows it is likely a Memory-to-Memory Vector + Architecture, and overcame the penalties normally associated + with this by adding an explicit "Vector operand forwarding/chaining" + instruction (Page 3-69). It is however clearly Scalable, up to Vector + elements of 2^16. The term Horizontal or Vertical alludes to the Matrix "Row-First" or "Column-First" technique, where: -- 2.30.2