2026-09-22T15:41:17.092595Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:compile_runtime: close time.busy=276µs time.idle=6.05µs
2026-09-22T15:42:22.451151Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:construct_mid:gemm_candidates: close time.busy=65.4s time.idle=3.03µs
2026-09-22T15:44:07.593385Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:construct_mid:gemm_candidates: close time.busy=105s time.idle=10.8µs
2026-09-22T15:44:13.893039Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:construct_mid: close time.busy=177s time.idle=3.92µs
2026-09-22T15:44:14.003765Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: wrote planner memory profile path=artifacts/kernel-cost-mlp-20260922/fp8-b4-n1-default/memory/baseline-1176495-0000.json scope="baseline" peak_step=2 effective_peak_bytes=320512
2026-09-22T15:44:15.480428Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:group_exchanges: close time.busy=130µs time.idle=3.69µs
2026-09-22T15:44:15.482525Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:merge_copies: close time.busy=1.79ms time.idle=1.22µs
2026-09-22T15:44:15.526324Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:eliminate: eliminated padding clears using kernel access contracts removed=0
2026-09-22T15:44:15.528066Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:eliminate: close time.busy=45.5ms time.idle=1.37µs
2026-09-22T15:44:15.583939Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:place_provisional_storage: close time.busy=55.1ms time.idle=2.91µs
2026-09-22T15:44:16.023933Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=1 transfers=27312 maximum_identical_destinations=1
2026-09-22T15:44:16.023959Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=1 initial_horizon=15233 selected_horizon=15233 endpoint_lower_bound=13176 lower_bound_gap=2057 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:44:16.148119Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=4 transfers=2304 maximum_identical_destinations=1
2026-09-22T15:44:16.148147Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=4 initial_horizon=21139 selected_horizon=21125 endpoint_lower_bound=20736 lower_bound_gap=389 selected_kind="critical-neighborhood" neighborhood_improvements=1
2026-09-22T15:44:17.050159Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=209304
2026-09-22T15:44:19.384318Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=19852
2026-09-22T15:44:19.652670Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=384480
2026-09-22T15:44:19.786814Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: compared ordinary and paired exchange schedules phase=3 candidates=41292 ordinary_horizon=17903 paired_horizon=16538 use_paired=true
2026-09-22T15:44:19.859270Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=3 transfers=43740 maximum_identical_destinations=1
2026-09-22T15:44:19.859292Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=3 initial_horizon=16538 selected_horizon=16538 endpoint_lower_bound=9216 lower_bound_gap=7322 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:44:19.926887Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=209304 elapsed_ms=2876 success=true
2026-09-22T15:44:23.636883Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=6 transfers=209304 maximum_identical_destinations=1
2026-09-22T15:44:23.636916Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=6 initial_horizon=1663 selected_horizon=1663 endpoint_lower_bound=1296 lower_bound_gap=367 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:44:25.696538Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=384480 elapsed_ms=6043 success=true
2026-09-22T15:44:26.489228Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=19852 elapsed_ms=7104 success=true
2026-09-22T15:44:29.428444Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=766680
2026-09-22T15:44:32.646747Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=19852
2026-09-22T15:44:33.332723Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=5 transfers=384480 maximum_identical_destinations=1
2026-09-22T15:44:33.332752Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=5 initial_horizon=19831 selected_horizon=19831 endpoint_lower_bound=18144 lower_bound_gap=1687 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:44:37.156649Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=19852 elapsed_ms=4509 success=true
2026-09-22T15:44:38.893053Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: compared ordinary and paired exchange schedules phase=0 candidates=7412 ordinary_horizon=50443 paired_horizon=50750 use_paired=false
2026-09-22T15:44:39.054322Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=0 transfers=19852 maximum_identical_destinations=1
2026-09-22T15:44:39.054344Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=0 initial_horizon=50443 selected_horizon=50443 endpoint_lower_bound=47328 lower_bound_gap=3115 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:44:47.757832Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=766680 elapsed_ms=18329 success=true
2026-09-22T15:45:08.852458Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=2 transfers=766680 maximum_identical_destinations=1
2026-09-22T15:45:08.852493Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=2 initial_horizon=6715 selected_horizon=6715 endpoint_lower_bound=4392 lower_bound_gap=2323 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:45:10.301689Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: close time.busy=54.7s time.idle=5.06µs
2026-09-22T15:45:10.360302Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:plan_kernels: close time.busy=58.5ms time.idle=3.13µs
2026-09-22T15:45:10.361978Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:compile_kernels: close time.busy=1.66ms time.idle=1.09µs
2026-09-22T15:45:10.362289Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:link_runtime: close time.busy=219µs time.idle=1.56µs
2026-09-22T15:45:10.749268Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: exchange table and repeat patch storage exchange_table_bytes=23956 maximum_uncompressed_repeat_patch_bytes=0 maximum_elided_arithmetic_patch_bytes=0
2026-09-22T15:45:11.262864Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:size_tile_code: close time.busy=32.5ms time.idle=5.99µs
2026-09-22T15:45:11.262927Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: placing generated tile programs linked_end=319744 host_code_base=376832 host_code_bytes=748 generated_code_bytes=1232 exchange_table_bytes=23956 executable_ranges=[(377648, 524288)]
2026-09-22T15:45:11.262985Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319744 profile_bytes=96 exchange_table_bytes=23956 host_code_bytes=748 host_data_bytes=528 generated_code_bytes=1232 code_address=377648 available_ranges=[(327680, 360448), (362736, 376832), (425984, 949168)]
2026-09-22T15:45:11.356256Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:place_storage: close time.busy=74.8ms time.idle=3.07µs
2026-09-22T15:45:11.613549Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=4
2026-09-22T15:45:11.625011Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=4 transfers=2304 maximum_identical_destinations=1
2026-09-22T15:45:11.625027Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=4 initial_horizon=21125 selected_horizon=21125 endpoint_lower_bound=20736 lower_bound_gap=389 selected_kind="reused" neighborhood_improvements=0
2026-09-22T15:45:11.625551Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=1
2026-09-22T15:45:11.633342Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=1 transfers=27312 maximum_identical_destinations=1
2026-09-22T15:45:11.633357Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=1 initial_horizon=15233 selected_horizon=15233 endpoint_lower_bound=13176 lower_bound_gap=2057 selected_kind="reused" neighborhood_improvements=0
2026-09-22T15:45:12.657947Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=3
2026-09-22T15:45:12.699703Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=3 transfers=43740 maximum_identical_destinations=1
2026-09-22T15:45:12.699726Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=3 initial_horizon=16538 selected_horizon=16538 endpoint_lower_bound=9216 lower_bound_gap=7322 selected_kind="reused" neighborhood_improvements=0
2026-09-22T15:45:17.646705Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=0
2026-09-22T15:45:17.781518Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=0 transfers=19852 maximum_identical_destinations=1
2026-09-22T15:45:17.781544Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=0 initial_horizon=50443 selected_horizon=50443 endpoint_lower_bound=47328 lower_bound_gap=3115 selected_kind="reused" neighborhood_improvements=0
2026-09-22T15:45:18.192906Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=209304
2026-09-22T15:45:20.464184Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=5
2026-09-22T15:45:20.600364Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=5 transfers=384480 maximum_identical_destinations=1
2026-09-22T15:45:20.600390Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=5 initial_horizon=19831 selected_horizon=19831 endpoint_lower_bound=18144 lower_bound_gap=1687 selected_kind="reused" neighborhood_improvements=0
2026-09-22T15:45:21.799889Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: finished incremental instruction-alignment validation transfers=209304 elapsed_ms=3606 success=true
2026-09-22T15:45:25.670422Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=6 transfers=209304 maximum_identical_destinations=1
2026-09-22T15:45:25.670456Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=6 initial_horizon=1663 selected_horizon=1663 endpoint_lower_bound=1296 lower_bound_gap=367 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T15:45:36.916978Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=2
2026-09-22T15:45:37.212971Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=2 transfers=766680 maximum_identical_destinations=1
2026-09-22T15:45:37.213018Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=2 initial_horizon=6715 selected_horizon=6715 endpoint_lower_bound=4392 lower_bound_gap=2323 selected_kind="reused" neighborhood_improvements=0
2026-09-22T15:45:38.847915Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: close time.busy=27.5s time.idle=6.01µs
2026-09-22T15:45:38.851745Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: costed final placed program final_cycles=584994 final_exchange=135748
2026-09-22T15:45:39.341912Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:prepare_tile_code: close time.busy=62.0ms time.idle=11.4µs
2026-09-22T15:45:39.381161Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:emit_tile_code: close time.busy=39.2ms time.idle=4.12µs
2026-09-22T15:45:39.578129Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:build_tile_images: close time.busy=188ms time.idle=2.90µs
2026-09-22T15:45:39.824717Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: wrote exact tile placement profile path=artifacts/kernel-cost-mlp-20260922/fp8-b4-n1-default/memory/placement-1176495.html
2026-09-22T15:45:39.968790Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: close time.busy=263s time.idle=7.25µs
2026-09-22T15:45:39.974772Z  INFO writing application package tiles=1472
2026-09-22T15:45:40.353300Z  INFO application package written
2026-09-22T15:45:40.353368Z  INFO reading application package
2026-09-22T15:45:40.514828Z  INFO application package read tiles=1472 compiler=0.1.0
2026-09-22T15:45:40.514916Z  INFO waiting for hardware ownership path=/tmp/ipu-stack-device.lock
2026-09-22T15:45:40.514925Z  INFO acquired hardware ownership path=/tmp/ipu-stack-device.lock
2026-09-22T15:45:40.514931Z  INFO opening IPU device path="/dev/ipu0"
2026-09-22T15:45:40.514979Z  INFO IPU device attached path=/dev/ipu0
2026-09-22T15:45:40.514982Z  INFO initializing IPU device
2026-09-22T15:45:40.523451Z  INFO IPU device initialized
2026-09-22T15:45:40.523515Z  INFO replaying device configuration register_writes=2003
2026-09-22T15:45:40.530337Z  INFO loading application tile_count=1472 final_mark=23
2026-09-22T15:45:40.754895Z  INFO application loaded and initial synchronization completed tile_count=1472
2026-09-22T15:45:41.681015Z  INFO prepared reference inputs elapsed_ms=922
2026-09-22T15:45:42.181627Z  INFO configured host reference BLAS threads=4
referencePrecision=device
2026-09-22T15:45:43.602864Z  INFO evaluated host reference elapsed_ms=1921
2026-09-22T15:45:43.606185Z  INFO starting host exchange session startup_mark=23
2026-09-22T15:45:43.606241Z  INFO attaching host exchange pages pages=2
2026-09-22T15:45:43.606987Z  INFO host exchange session started
2026-09-22T15:45:43.607018Z  INFO invoking host exchange call call="initialize" command=0 phases=4
2026-09-22T15:45:43.619084Z  INFO invoking host exchange call call="run" command=0 phases=15
referenceInference=1 numericalTest=PASS
profile=artifacts/kernel-cost-mlp-20260922/fp8-b4-n1-default/profile.capnp tiles=1472 samples=27117
workload=siglip-mlp-f16-b4-t729-d1152-h4304-n1 benchmark=siglip-mlp-f16 batch=4 tokens=729 rows=2916 dimension=1152 hiddenDimension=4304 blocks=1 biases=false activeTiles=1458 executionTiles=1472 inputBytes=3359232 weightBytes=9934848 cycles=577326 minimumTileCycles=569988 deviceMicroseconds=384.884 effectiveGemmTflops=150.260 peakTflops=282.624 efficiencyPercent=53.17 maximumAbsoluteError=0.006104
package=artifacts/kernel-cost-mlp-20260922/fp8-b4-n1-default/model.ipuexe tiles=1472 hardwareTest=PASS