2026-09-22T14:51:10.284072Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:compile_runtime: close time.busy=342µs time.idle=5.75µs
2026-09-22T14:52:45.905498Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:construct_mid:gemm_candidates: close time.busy=95.6s time.idle=3.06µs
2026-09-22T14:55:57.298579Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:construct_mid:gemm_candidates: close time.busy=191s time.idle=3.58µs
2026-09-22T14:56:01.402675Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:construct_mid: close time.busy=291s time.idle=3.67µs
2026-09-22T14:56:01.494769Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: wrote planner memory profile path=artifacts/planner-memory-20260922/fp8-b4-n1-default/memory/baseline-1170924-0000.json scope="baseline" peak_step=2 effective_peak_bytes=320512
2026-09-22T14:56:02.504452Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:group_exchanges: close time.busy=105µs time.idle=3.15µs
2026-09-22T14:56:02.505973Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:merge_copies: close time.busy=1.21ms time.idle=765ns
2026-09-22T14:56:02.537169Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:eliminate: eliminated padding clears using kernel access contracts removed=0
2026-09-22T14:56:02.538433Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:eliminate: close time.busy=32.4ms time.idle=790ns
2026-09-22T14:56:02.581905Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:place_provisional_storage: close time.busy=42.9ms time.idle=1.27µs
2026-09-22T14:56:02.939779Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=4 transfers=1728 maximum_identical_destinations=1
2026-09-22T14:56:02.939811Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=4 initial_horizon=21814 selected_horizon=21802 endpoint_lower_bound=21504 lower_bound_gap=298 selected_kind="critical-neighborhood" neighborhood_improvements=1
2026-09-22T14:56:03.058092Z  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
2026-09-22T14:56:03.058119Z  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
2026-09-22T14:56:04.077421Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=58320
2026-09-22T14:56:04.545764Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=209448
2026-09-22T14:56:05.915527Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=19852
2026-09-22T14:56:06.147492Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=58320 elapsed_ms=2070 success=true
2026-09-22T14:56:08.064498Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=209448 elapsed_ms=3518 success=true
2026-09-22T14:56:08.512650Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=58320
2026-09-22T14:56:10.829807Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=6 transfers=209448 maximum_identical_destinations=1
2026-09-22T14:56:10.829845Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=6 initial_horizon=1672 selected_horizon=1672 endpoint_lower_bound=1344 lower_bound_gap=328 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T14:56:10.865352Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=58320 elapsed_ms=2352 success=true
2026-09-22T14:56:11.809382Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: compared ordinary and paired exchange schedules phase=3 candidates=816 ordinary_horizon=14660 paired_horizon=14886 use_paired=false
2026-09-22T14:56:11.874550Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=19852 elapsed_ms=5958 success=true
2026-09-22T14:56:11.886748Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=3 transfers=58320 maximum_identical_destinations=1
2026-09-22T14:56:11.886764Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=3 initial_horizon=14660 selected_horizon=14660 endpoint_lower_bound=13664 lower_bound_gap=996 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T14:56:13.290681Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=685728
2026-09-22T14:56:17.361156Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=19852
2026-09-22T14:56:19.255641Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=766680
2026-09-22T14:56:24.256574Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=19852 elapsed_ms=6895 success=true
2026-09-22T14:56:26.270251Z  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
2026-09-22T14:56:26.456338Z  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
2026-09-22T14:56:26.456360Z  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
2026-09-22T14:56:28.962603Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=685728 elapsed_ms=15671 success=true
2026-09-22T14:56:43.290787Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: finished incremental instruction-alignment validation transfers=766680 elapsed_ms=24035 success=true
2026-09-22T14:56:46.507773Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: prepared large physical exchange phase phase=5 transfers=685728 maximum_identical_destinations=1
2026-09-22T14:56:46.507803Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: optimized physical exchange schedule phase=5 initial_horizon=28630 selected_horizon=28630 endpoint_lower_bound=25536 lower_bound_gap=3094 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T14:57:01.945228Z  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
2026-09-22T14:57:01.945259Z  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
2026-09-22T14:57:03.374603Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:schedule_provisional_exchanges: close time.busy=60.8s time.idle=3.13µs
2026-09-22T14:57:03.401650Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:plan_kernels: close time.busy=27.0ms time.idle=2.22µs
2026-09-22T14:57:03.403750Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:compile_kernels: close time.busy=2.08ms time.idle=1.53µs
2026-09-22T14:57:03.403919Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:link_runtime: close time.busy=115µs time.idle=576ns
2026-09-22T14:57:03.662625Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: exchange table and repeat patch storage exchange_table_bytes=28444 maximum_uncompressed_repeat_patch_bytes=0 maximum_elided_arithmetic_patch_bytes=0
2026-09-22T14:57:04.064660Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:size_tile_code: close time.busy=62.2ms time.idle=4.13µs
2026-09-22T14:57:04.064692Z  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=1804 exchange_table_bytes=28444 executable_ranges=[(377648, 524288)]
2026-09-22T14:57:04.064724Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319744 profile_bytes=192 exchange_table_bytes=28444 host_code_bytes=748 host_data_bytes=528 generated_code_bytes=1804 code_address=377648 available_ranges=[(327680, 360448), (362736, 376832), (425984, 949168)]
2026-09-22T14:57:04.131465Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:place_storage: close time.busy=55.8ms time.idle=2.22µs
2026-09-22T14:57:04.266379Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=1
2026-09-22T14:57:04.271452Z  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
2026-09-22T14:57:04.271462Z  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
2026-09-22T14:57:04.400868Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=4
2026-09-22T14:57:04.408798Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=4 transfers=1728 maximum_identical_destinations=1
2026-09-22T14:57:04.408811Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=4 initial_horizon=21802 selected_horizon=21802 endpoint_lower_bound=21504 lower_bound_gap=298 selected_kind="reused" neighborhood_improvements=0
2026-09-22T14:57:06.452461Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=3
2026-09-22T14:57:06.500086Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=3 transfers=58320 maximum_identical_destinations=1
2026-09-22T14:57:06.500118Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=3 initial_horizon=14660 selected_horizon=14660 endpoint_lower_bound=13664 lower_bound_gap=996 selected_kind="reused" neighborhood_improvements=0
2026-09-22T14:57:09.254574Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=209448
2026-09-22T14:57:10.427053Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=0
2026-09-22T14:57:10.511982Z  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
2026-09-22T14:57:10.512003Z  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
2026-09-22T14:57:11.555247Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: finished incremental instruction-alignment validation transfers=209448 elapsed_ms=2300 success=true
2026-09-22T14:57:14.195049Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=6 transfers=209448 maximum_identical_destinations=1
2026-09-22T14:57:14.195085Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=6 initial_horizon=1672 selected_horizon=1672 endpoint_lower_bound=1344 lower_bound_gap=328 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T14:57:22.896346Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: reused exchange optimization after validating relocated rows phase=2
2026-09-22T14:57:23.168296Z  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
2026-09-22T14:57:23.168321Z  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
2026-09-22T14:57:27.359804Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: retrying exchange schedule with incremental instruction-alignment validation transfers=685728
2026-09-22T14:57:41.076135Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: finished incremental instruction-alignment validation transfers=685728 elapsed_ms=13716 success=true
2026-09-22T14:57:56.006397Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: prepared large physical exchange phase phase=5 transfers=685728 maximum_identical_destinations=1
2026-09-22T14:57:56.006425Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: optimized physical exchange schedule phase=5 initial_horizon=28426 selected_horizon=28426 endpoint_lower_bound=25536 lower_bound_gap=2890 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-22T14:57:57.120795Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:lower_exchanges: close time.busy=53.0s time.idle=3.10µs
2026-09-22T14:57:57.123654Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: costed final placed program final_cycles=761840 final_exchange=143151
2026-09-22T14:57:57.779672Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:prepare_tile_code: close time.busy=150ms time.idle=6.28µs
2026-09-22T14:57:57.802471Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:emit_tile_code: close time.busy=22.7ms time.idle=2.50µs
2026-09-22T14:57:57.971552Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:build_tile_images: close time.busy=163ms time.idle=1.81µs
2026-09-22T14:57:58.198174Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: wrote exact tile placement profile path=artifacts/planner-memory-20260922/fp8-b4-n1-default/memory/placement-1170924.html
2026-09-22T14:57:58.305298Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: close time.busy=408s time.idle=4.95µs
2026-09-22T14:57:58.309152Z  INFO writing application package tiles=1472
2026-09-22T14:57:58.460076Z  INFO application package written
2026-09-22T14:57:58.460120Z  INFO reading application package
2026-09-22T14:57:58.698801Z  INFO application package read tiles=1472 compiler=0.1.0
2026-09-22T14:57:58.698889Z  INFO waiting for hardware ownership path=/tmp/ipu-stack-device.lock
2026-09-22T14:57:58.698899Z  INFO acquired hardware ownership path=/tmp/ipu-stack-device.lock
2026-09-22T14:57:58.698912Z  INFO opening IPU device path="/dev/ipu0"
2026-09-22T14:57:58.698948Z  INFO IPU device attached path=/dev/ipu0
2026-09-22T14:57:58.698953Z  INFO initializing IPU device
2026-09-22T14:57:58.706733Z  INFO IPU device initialized
2026-09-22T14:57:58.706747Z  INFO replaying device configuration register_writes=2003
2026-09-22T14:57:58.714930Z  INFO loading application tile_count=1472 final_mark=23
2026-09-22T14:57:58.912608Z  INFO application loaded and initial synchronization completed tile_count=1472
2026-09-22T14:57:59.815138Z  INFO prepared reference inputs elapsed_ms=899
2026-09-22T14:58:00.294368Z  INFO configured host reference BLAS threads=4
referencePrecision=device
2026-09-22T14:58:01.505759Z  INFO evaluated host reference elapsed_ms=1690
2026-09-22T14:58:01.507843Z  INFO starting host exchange session startup_mark=23
2026-09-22T14:58:01.507859Z  INFO attaching host exchange pages pages=2
2026-09-22T14:58:01.508531Z  INFO host exchange session started
2026-09-22T14:58:01.508554Z  INFO invoking host exchange call call="initialize" command=0 phases=4
2026-09-22T14:58:01.511408Z  INFO invoking host exchange call call="run" command=0 phases=15
referenceInference=1 numericalTest=PASS
profile=artifacts/planner-memory-20260922/fp8-b4-n1-default/profile.capnp tiles=1472 samples=28825
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=614700 minimumTileCycles=607116 deviceMicroseconds=409.800 effectiveGemmTflops=141.124 peakTflops=282.624 efficiencyPercent=49.93 maximumAbsoluteError=0.006470
package=artifacts/planner-memory-20260922/fp8-b4-n1-default/model.ipuexe tiles=1472 hardwareTest=PASS