code wiki / (root) / nx_simd_i32x8_test.nx

nx_simd_i32x8_test.nx

buildroot/runtime/nx_simd_i32x8_test.nx

4729 B130 linesdepth 6pulls 8 transitivereach 0 importersview sourcekind gate/prooftopic simd
docsdependenciesstructsconstsfunctions

about

nx_simd_i32x8_test.nx -- exercise the i32x8 SIMD intrinsics. Same scalar-vs-SIMD-bit-exact gate as nx_simd_test.nx but with 8 i32 lanes per vector (double the throughput of i64x4 for 32-bit workloads -- ML token IDs, image pixels, hashes).

dependencies 2 imports · 0 importers

nx_kernel_v2.nx nx_log.nx nx_simd_i32x8_test.nx

imports: nx_kernel_v2.nxnx_log.nx

imported by: nobody (leaf or entry point)

call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown

main println sys_write strlen sys_mmap nxa_die sys_write ↻ sys_exit nxa_lock_take nxa_lock_addr sys_write ↻ nxa_lock_give nxa_lock_addr ↻ nxa_report_overrun sys_write ↻ nxa_dump_printable sys_write ↻ nxa_dump_sizes sys_write ↻ t1_round_trip_load_store t2_vadd t3_vmul t4_vbroadcast t5_vreduce_sum

structs

none

consts

none

functions

14func t1_round_trip_load_store() -> i64
called by 1: main
34func t2_vadd() -> i64
called by 1: main
62func t3_vmul() -> i64
called by 1: main
89func t4_vbroadcast() -> i64
called by 1: main
101func t5_vreduce_sum() -> i64
called by 1: main
115func main() -> nx_exit