Files
FFmpeg/libavutil
Jeongkeun Kim de18feb0f0 avutil/aarch64: add pixelutils 32x32 SAD NEON implementation
This adds a NEON-optimized function for computing 32x32 Sum of Absolute
Differences (SAD) on AArch64, addressing a gap where x86 had SSE2/AVX2
implementations but AArch64 lacked equivalent coverage.

The implementation mirrors the existing sad8 and sad16 NEON functions,
employing a 4-row unrolled loop with UABAL and UABAL2 instructions for
efficient load-compute interleaving, and four 8x16-bit accumulators to
handle the wider 32-byte rows.

Benchmarks on AWS Graviton3 (Neoverse V1, c7g.xlarge) using checkasm:
  sad_32x32_0: C 146.4 cycles -> NEON  98.1 cycles (1.49x speedup)
  sad_32x32_1: C 141.4 cycles -> NEON  98.9 cycles (1.43x speedup)
  sad_32x32_2: C 140.7 cycles -> NEON  95.0 cycles (1.48x speedup)

Signed-off-by: Jeongkeun Kim <variety0724@gmail.com>
2026-04-19 19:27:55 +00:00
..
2025-08-03 13:48:47 +02:00
2021-04-27 10:43:13 -03:00
2024-07-21 17:02:12 +02:00
2025-05-07 15:35:47 +02:00
2024-03-04 17:45:00 +01:00
2025-08-03 13:48:47 +02:00
2014-07-12 22:33:27 +02:00
2026-03-02 11:55:30 +01:00
2023-11-18 19:55:42 +01:00
2024-03-15 12:51:15 +01:00
2025-09-02 17:05:50 +02:00
2021-07-22 14:34:31 +02:00
2021-07-22 14:34:31 +02:00
2026-01-18 03:33:04 +00:00
2025-08-03 13:48:47 +02:00
2025-12-23 03:39:16 +00:00
2015-10-11 18:03:10 -04:00
2025-08-03 13:48:47 +02:00
2024-06-01 18:05:58 +03:00
2025-08-03 13:48:47 +02:00
2021-07-21 16:35:27 +02:00
2023-09-07 00:42:10 +02:00
2021-07-22 14:34:31 +02:00
2023-09-07 00:42:10 +02:00
2023-09-07 00:42:10 +02:00
2021-07-22 14:34:31 +02:00
2025-08-03 13:48:47 +02:00
2024-03-31 00:08:43 +01:00
2025-09-02 17:05:50 +02:00
2025-08-03 13:48:47 +02:00
2025-08-03 13:48:47 +02:00
2025-08-03 13:48:47 +02:00
2021-07-22 14:34:31 +02:00
2024-01-16 01:34:57 +01:00
2021-07-22 14:34:31 +02:00
2021-07-22 14:34:31 +02:00
2025-08-03 13:48:47 +02:00
2021-07-22 14:34:31 +02:00
2022-10-17 09:51:47 +02:00
2023-09-07 00:42:10 +02:00
2025-08-03 13:48:47 +02:00
2023-08-08 09:46:11 +02:00
2025-08-03 13:48:47 +02:00