From 75b7359983967f6866987611e0793df5d6f0ecda Mon Sep 17 00:00:00 2001 From: lkcl Date: Fri, 22 Oct 2021 17:58:32 +0100 Subject: [PATCH] --- 3d_gpu/architecture/dynamic_simd/simdscope.mdwn | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/3d_gpu/architecture/dynamic_simd/simdscope.mdwn b/3d_gpu/architecture/dynamic_simd/simdscope.mdwn index 720d21bb5..d7db5bd62 100644 --- a/3d_gpu/architecture/dynamic_simd/simdscope.mdwn +++ b/3d_gpu/architecture/dynamic_simd/simdscope.mdwn @@ -59,8 +59,9 @@ Here, because SimdShape derives from Shape, things still work because SimdShape works out that its maximum scalar size is 11, and sets Shape.width to 11 when SimdScope is set in scalar mode. When scalar=False, full SIMD is activated and -the resultant HDL combines vec_el_counts with expshape to create -1xFP64, 2xFP32, 4xFP16, 4xBF16 where the exponents are 11, 8, 5, 8 +the resultant HDL combines vec_el_counts with expshape as part +of an ALU with inputs that can be +1xFP64, 2xFP32, 4xFP16, 4xBF16, where the exponents are 1x11, 2x8, 4x5, 4x8 respectively. Behind the scenes, when calling SimdShape.Signal in SIMD mode, -- 2.30.2