ffmpeg/libavcodec/aarch64/vvc
Georgii Zagoruiko 1ced59326a aarch64/vvc: Optimisations of put_chroma_hv() functions for 10/12-bit
Apple M4:
put_chroma_hv_10_2x2_c:                                  9.1 ( 1.00x)
put_chroma_hv_10_4x4_c:                                 20.1 ( 1.00x)
put_chroma_hv_10_8x8_c:                                 35.6 ( 1.00x)
put_chroma_hv_10_8x8_neon:                              15.4 ( 2.31x)
put_chroma_hv_10_16x16_c:                              113.7 ( 1.00x)
put_chroma_hv_10_16x16_neon:                            57.0 ( 1.99x)
put_chroma_hv_10_32x32_c:                              406.9 ( 1.00x)
put_chroma_hv_10_32x32_neon:                           225.7 ( 1.80x)
put_chroma_hv_10_64x64_c:                             1498.8 ( 1.00x)
put_chroma_hv_10_64x64_neon:                           876.2 ( 1.71x)
put_chroma_hv_10_128x128_c:                           5757.0 ( 1.00x)
put_chroma_hv_10_128x128_neon:                        3446.6 ( 1.67x)
put_chroma_hv_12_2x2_c:                                  9.9 ( 1.00x)
put_chroma_hv_12_4x4_c:                                 19.2 ( 1.00x)
put_chroma_hv_12_8x8_c:                                 36.1 ( 1.00x)
put_chroma_hv_12_8x8_neon:                              17.9 ( 2.02x)
put_chroma_hv_12_16x16_c:                              112.2 ( 1.00x)
put_chroma_hv_12_16x16_neon:                            55.6 ( 2.02x)
put_chroma_hv_12_32x32_c:                              416.6 ( 1.00x)
put_chroma_hv_12_32x32_neon:                           224.3 ( 1.86x)
put_chroma_hv_12_64x64_c:                             1464.8 ( 1.00x)
put_chroma_hv_12_64x64_neon:                           860.1 ( 1.70x)
put_chroma_hv_12_128x128_c:                           5776.8 ( 1.00x)
put_chroma_hv_12_128x128_neon:                        3445.2 ( 1.68x)

RPi5:
put_chroma_hv_10_2x2_c:                                118.5 ( 1.00x)
put_chroma_hv_10_4x4_c:                                190.6 ( 1.00x)
put_chroma_hv_10_8x8_c:                                303.1 ( 1.00x)
put_chroma_hv_10_8x8_neon:                             172.6 ( 1.76x)
put_chroma_hv_10_16x16_c:                             1036.1 ( 1.00x)
put_chroma_hv_10_16x16_neon:                           626.7 ( 1.65x)
put_chroma_hv_10_32x32_c:                             3624.4 ( 1.00x)
put_chroma_hv_10_32x32_neon:                          2386.9 ( 1.52x)
put_chroma_hv_10_64x64_c:                            13612.1 ( 1.00x)
put_chroma_hv_10_64x64_neon:                          9314.8 ( 1.46x)
put_chroma_hv_10_128x128_c:                          52975.4 ( 1.00x)
put_chroma_hv_10_128x128_neon:                       37083.5 ( 1.43x)
put_chroma_hv_12_2x2_c:                                118.6 ( 1.00x)
put_chroma_hv_12_4x4_c:                                188.1 ( 1.00x)
put_chroma_hv_12_8x8_c:                                303.4 ( 1.00x)
put_chroma_hv_12_8x8_neon:                             176.7 ( 1.72x)
put_chroma_hv_12_16x16_c:                             1037.9 ( 1.00x)
put_chroma_hv_12_16x16_neon:                           626.5 ( 1.66x)
put_chroma_hv_12_32x32_c:                             3629.0 ( 1.00x)
put_chroma_hv_12_32x32_neon:                          2386.6 ( 1.52x)
put_chroma_hv_12_64x64_c:                            13649.0 ( 1.00x)
put_chroma_hv_12_64x64_neon:                          9313.6 ( 1.47x)
put_chroma_hv_12_128x128_c:                          52978.0 ( 1.00x)
put_chroma_hv_12_128x128_neon:                       37101.2 ( 1.43x)
2026-04-27 20:10:57 +00:00
..
alf.S avcodec/aarch64/vvc: Optimised version of classify function. 2025-09-09 22:13:04 +01:00
alf_template.c aarch64/vvc: Fix DCE undefined references with MSVC 2026-03-05 11:57:40 +02:00
dsp_init.c aarch64/vvc: Optimisations of put_chroma_hv() functions for 10/12-bit 2026-04-27 20:10:57 +00:00
inter.S aarch64/vvc: Optimisations of put_chroma_hv() functions for 10/12-bit 2026-04-27 20:10:57 +00:00
inter_sme2.S aarch64/inter_sme2: Remove needless backup/restore of x29/x30 2026-03-20 13:16:06 +02:00
Makefile aarch64/vvc: sme2 optimisation of alf_filter_luma() 8/10/12 bit 2026-03-04 23:52:58 +02:00
sad.S aarch64/vvc: Add sad 2024-09-14 16:36:34 +08:00