From: lkcl Date: Sat, 9 Oct 2021 14:43:15 +0000 (+0100) Subject: (no commit message) X-Git-Tag: opf_rfc_ls005_v1~3674 X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=7c69d4e0eef77e12c050cd1c627c5781dc386f8b;p=libreriscv.git --- diff --git a/3d_gpu/architecture/dynamic_simd/shape.mdwn b/3d_gpu/architecture/dynamic_simd/shape.mdwn index 3fdc4a3d7..1f64ebc9f 100644 --- a/3d_gpu/architecture/dynamic_simd/shape.mdwn +++ b/3d_gpu/architecture/dynamic_simd/shape.mdwn @@ -67,7 +67,7 @@ and that "x" marks unused (padding) portions: thus, we deduce, we *actually* need breakpoints at these positions, and that unused portions common to **all** cases can be deduced -and marked "x" +and marked "x" by looking at the columns above which all contain "x": | |28|26|24| |20|16| |12|10|8| |4 | x x @@ -78,7 +78,7 @@ these sub-portions it is unnecessary to allocate computational gates. Also in order to save gates, in the example above there are only three cases (32 bit, 16 bit, 8 bit) therefore only three sets of logic are required to construct the larger overall computational result -from the "smaller chunks", rather than at first glance, with there +from the "smaller chunks". At first glance, with there being 9 actual partitions (28, 26, 24, 20, 16, 12, 10, 8, 4), it would appear that 2^9 (512!) cases were required, where in fact there are only three.