date: 2026-09-15T09:39:17Z os: Darwin 25.2.0 arm64 macos: 26.2 (25C56) cpu: Apple M4 Pro logical cpus: 14 performance cores: 10 efficiency cores: 4 smt: none P-core L1d: 131072 B L2 (shared per cluster of 5): 16777216 B E-core L1d: 65536 B L2 (shared per cluster of 4): 4194304 B cache line: 128 B page: 16384 B memory: 24 GiB isa features: CRC32 FlagM FlagM2 FHM DotProd SHA3 RDM LSE SHA256 SHA512 SHA1 AES PMULL SB FRINTTS PACIMP LRCPC LRCPC2 FCMA JSCVT PAuth PAuth2 FPAC FPACCOMBINE DPB DPB2 BF16 I8MM WFxT RPRES ECV AFP LSE2 CSV2 CSV3 DIT FP16 BTI SME SME2 SME_F64F64 SME_I16I64 power: 'AC Power' load average at start: 10.21 8.24 8.04 frequency: not published by the vendor for this part; see estimated clock below compiler: Apple clang version 17.0.0 (clang-1700.4.4.1) kernel: Darwin Kernel Version 25.2.0: Tue Nov 18 21:09:56 PST 2025; root:xnu-12377.61.12~1/RELEASE_ARM64_T6041 boot args: none set mitigation state: not exposed by macOS estimated clock: 4.49 GHz (dependent 1-cycle add chain, 400000000 adds, min of 7 runs) bench: QUICK=0 REPS=default CLOCK_GHZ=4.49 --- file /var/folders/t4/dv716g7154qc59jxlrdxyn_h0000gn/T//syscall-cost.wbGZy7/data bytes 67108864 written and fsynced in 46 ms, read back and verified in 5 ms read offset 33554432 reps 31 (plus 1 warmup, discarded) clock 4.49 GHz size 1 bytes: 65536 calls per rep size 64 bytes: 65536 calls per rep size 4096 bytes: 65536 calls per rep size 65536 bytes: 4096 calls per rep size 1048576 bytes: 256 calls per rep RESULT clock 4.49 GHz RESULT reps 31 runs pread 1 B min 281.246 median 290.455 max 328.547 cv 3.1% n 31 ns/call RESULT pread_1_bytes 1 bytes RESULT pread_1_calls 65536 calls/rep RESULT pread_1_min 281.246 ns/call RESULT pread_1_median 290.455 ns/call RESULT pread_1_cv 3.09 percent RESULT pread_1_nsbyte 281.24556 ns/byte RESULT pread_1_cycles 1262.8 cycles/call pread 64 B min 281.916 median 290.424 max 305.824 cv 2.0% n 31 ns/call RESULT pread_64_bytes 64 bytes RESULT pread_64_calls 65536 calls/rep RESULT pread_64_min 281.916 ns/call RESULT pread_64_median 290.424 ns/call RESULT pread_64_cv 1.96 percent RESULT pread_64_nsbyte 4.40493 ns/byte RESULT pread_64_cycles 1265.8 cycles/call pread 4096 B min 313.670 median 320.944 max 342.719 cv 1.9% n 31 ns/call RESULT pread_4096_bytes 4096 bytes RESULT pread_4096_calls 65536 calls/rep RESULT pread_4096_min 313.670 ns/call RESULT pread_4096_median 320.944 ns/call RESULT pread_4096_cv 1.91 percent RESULT pread_4096_nsbyte 0.07658 ns/byte RESULT pread_4096_cycles 1408.4 cycles/call pread 65536 B min 1257.192 median 1370.860 max 1527.669 cv 3.4% n 31 ns/call RESULT pread_65536_bytes 65536 bytes RESULT pread_65536_calls 4096 calls/rep RESULT pread_65536_min 1257.192 ns/call RESULT pread_65536_median 1370.860 ns/call RESULT pread_65536_cv 3.36 percent RESULT pread_65536_nsbyte 0.01918 ns/byte RESULT pread_65536_cycles 5644.8 cycles/call pread 1048576 B min 20490.398 median 21189.293 max 22415.363 cv 1.5% n 31 ns/call RESULT pread_1048576_bytes 1048576 bytes RESULT pread_1048576_calls 256 calls/rep RESULT pread_1048576_min 20490.398 ns/call RESULT pread_1048576_median 21189.293 ns/call RESULT pread_1048576_cv 1.53 percent RESULT pread_1048576_nsbyte 0.01954 ns/byte RESULT pread_1048576_cycles 92001.9 cycles/call memcpy 1 B min 1.108 median 1.209 max 1.286 cv 5.0% n 31 ns/call RESULT memcpy_1_bytes 1 bytes RESULT memcpy_1_calls 65536 calls/rep RESULT memcpy_1_min 1.108 ns/call RESULT memcpy_1_median 1.209 ns/call RESULT memcpy_1_cv 5.05 percent RESULT memcpy_1_nsbyte 1.10817 ns/byte RESULT memcpy_1_cycles 5.0 cycles/call memcpy 64 B min 1.330 median 1.429 max 2.118 cv 9.2% n 31 ns/call RESULT memcpy_64_bytes 64 bytes RESULT memcpy_64_calls 65536 calls/rep RESULT memcpy_64_min 1.330 ns/call RESULT memcpy_64_median 1.429 ns/call RESULT memcpy_64_cv 9.15 percent RESULT memcpy_64_nsbyte 0.02078 ns/byte RESULT memcpy_64_cycles 6.0 cycles/call memcpy 4096 B min 37.453 median 42.519 max 48.382 cv 5.2% n 31 ns/call RESULT memcpy_4096_bytes 4096 bytes RESULT memcpy_4096_calls 65536 calls/rep RESULT memcpy_4096_min 37.453 ns/call RESULT memcpy_4096_median 42.519 ns/call RESULT memcpy_4096_cv 5.17 percent RESULT memcpy_4096_nsbyte 0.00914 ns/byte RESULT memcpy_4096_cycles 168.2 cycles/call memcpy 65536 B min 519.897 median 546.021 max 599.436 cv 4.2% n 31 ns/call RESULT memcpy_65536_bytes 65536 bytes RESULT memcpy_65536_calls 4096 calls/rep RESULT memcpy_65536_min 519.897 ns/call RESULT memcpy_65536_median 546.021 ns/call RESULT memcpy_65536_cv 4.17 percent RESULT memcpy_65536_nsbyte 0.00793 ns/byte RESULT memcpy_65536_cycles 2334.3 cycles/call memcpy 1048576 B min 11197.266 median 12204.750 max 12880.859 cv 2.5% n 31 ns/call RESULT memcpy_1048576_bytes 1048576 bytes RESULT memcpy_1048576_calls 256 calls/rep RESULT memcpy_1048576_min 11197.266 ns/call RESULT memcpy_1048576_median 12204.750 ns/call RESULT memcpy_1048576_cv 2.53 percent RESULT memcpy_1048576_nsbyte 0.01068 ns/byte RESULT memcpy_1048576_cycles 50275.7 cycles/call getppid min 83.148 median 87.238 max 89.345 cv 1.6% n 31 ns/call RESULT getppid_bytes 0 bytes RESULT getppid_calls 65536 calls/rep RESULT getppid_min 83.148 ns/call RESULT getppid_median 87.238 ns/call RESULT getppid_cv 1.65 percent RESULT getppid_cycles 373.3 cycles/call clock_gettime MONOTONIC_RAW min 12.201 median 12.686 max 13.100 cv 2.2% n 31 ns/call RESULT clock_gettime_bytes 0 bytes RESULT clock_gettime_calls 65536 calls/rep RESULT clock_gettime_min 12.201 ns/call RESULT clock_gettime_median 12.686 ns/call RESULT clock_gettime_cv 2.24 percent RESULT clock_gettime_cycles 54.8 cycles/call pread: fixed cost 281.3 ns (intercept over 1, 64, 4096 B), slope 0.00790 ns/byte there; copy slope 0.01957 ns/byte between 65536 and 1048576 B memcpy: fixed cost 0.9 ns (intercept over 1, 64, 4096 B), slope 0.00892 ns/byte there; copy slope 0.01086 ns/byte between 65536 and 1048576 B RESULT pread_fixed 281.3 ns RESULT pread_small_slope 0.00790 ns/byte RESULT pread_copy_slope 0.01957 ns/byte RESULT memcpy_fixed 0.9 ns RESULT memcpy_small_slope 0.00892 ns/byte RESULT memcpy_copy_slope 0.01086 ns/byte RESULT pread_fixed_cycles 1263 cycles RESULT clock_checksum 1114345161822424 sum checksums: every pread and memcpy pass matched its reference sum and the source bytes; every getppid returned the parent pid