Re: [FFmpeg-devel] [PATCH] libavfilter/af_afir: R-V V dcmul_add

2023-12-22 Thread Rémi Denis-Courmont
Le perjantaina 22. joulukuuta 2023, 3.41.29 EET flow gg a écrit : > It's at c908 > > According to the benchmark results, if vlseg2e64 is used, the speed is > almost as slow as C language (dcmul_add_rvv_f64: 86.2), if vsseg2e64 is > used, it will be only a bit slower (dcmul_add_rvv_f64: 50.2). Fai

Re: [FFmpeg-devel] [PATCH] libavfilter/af_afir: R-V V dcmul_add

2023-12-21 Thread flow gg
It's at c908 According to the benchmark results, if vlseg2e64 is used, the speed is almost as slow as C language (dcmul_add_rvv_f64: 86.2), if vsseg2e64 is used, it will be only a bit slower (dcmul_add_rvv_f64: 50.2). Rémi Denis-Courmont 于2023年12月22日周五 04:52写道: > Le tiistaina 19. joulukuuta 202

Re: [FFmpeg-devel] [PATCH] libavfilter/af_afir: R-V V dcmul_add

2023-12-21 Thread Rémi Denis-Courmont
Le tiistaina 19. joulukuuta 2023, 4.53.12 EET flow gg a écrit : > c908: > dcmul_add_c: 88.0 > dcmul_add_rvv_f64: 46.2 > > Did not use vlseg2e64, because it is much slower than vlse64 > Did not use vsseg2e64, because it is slightly slower than vsse64 Is this about C910 or C908? I have not checked

[FFmpeg-devel] [PATCH] libavfilter/af_afir: R-V V dcmul_add

2023-12-18 Thread flow gg
c908: dcmul_add_c: 88.0 dcmul_add_rvv_f64: 46.2 Did not use vlseg2e64, because it is much slower than vlse64 Did not use vsseg2e64, because it is slightly slower than vsse64 From 80b6694bc29ed1c37852dc079a6d91a24dd6f18e Mon Sep 17 00:00:00 2001 From: sunyuechi Date: Tue, 19 Dec 2023 09:11:28 +080