2026-09-07T14:32:00.372689Z  INFO retained operator-plan finalist finalist=0 values=12 operations=9 estimated_cycles=335607 estimated_exchange_cycles=81520 exchange_row_bytes=17556 peak_standard_bytes=186004 peak_interleaved_bytes=79360 peak_total_bytes=232596 maximum_standard_allocation_bytes=163840 active_tile_counts={1440, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(0)), Some(OperationId(0)), Some(OperationId(1)), Some(OperationId(2)), Some(OperationId(2)), Some(OperationId(2))]
2026-09-07T14:32:00.381680Z  INFO retained operator-plan finalist finalist=1 values=11 operations=8 estimated_cycles=339304 estimated_exchange_cycles=106360 exchange_row_bytes=23276 peak_standard_bytes=191724 peak_interleaved_bytes=184320 peak_total_bytes=316652 maximum_standard_allocation_bytes=163840 active_tile_counts={1458, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(0)), Some(OperationId(0)), Some(OperationId(2)), Some(OperationId(2)), Some(OperationId(2))]
2026-09-07T14:32:00.389047Z  INFO retained operator-plan finalist finalist=2 values=12 operations=9 estimated_cycles=343357 estimated_exchange_cycles=98480 exchange_row_bytes=21372 peak_standard_bytes=189820 peak_interleaved_bytes=153600 peak_total_bytes=299388 maximum_standard_allocation_bytes=163840 active_tile_counts={1470, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(0)), Some(OperationId(0)), Some(OperationId(1)), Some(OperationId(2)), Some(OperationId(2)), Some(OperationId(2))]
2026-09-07T14:32:00.396534Z  INFO retained operator-plan finalist finalist=3 values=12 operations=9 estimated_cycles=346729 estimated_exchange_cycles=95440 exchange_row_bytes=20688 peak_standard_bytes=106704 peak_interleaved_bytes=83840 peak_total_bytes=186064 maximum_standard_allocation_bytes=81920 active_tile_counts={1440, 1449, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(0)), Some(OperationId(0)), Some(OperationId(1)), Some(OperationId(2)), Some(OperationId(2)), Some(OperationId(2))]
2026-09-07T14:32:00.500543Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_mid"}: package build phase finished phase="lower_mid" elapsed_ms=12219 success=true
2026-09-07T14:32:03.696848Z  INFO screened tile mappings by exchange resource load candidates=88 baseline_cycles=33786 candidate_cycles=33786 baseline_pressure=29866 candidate_pressure=29866
2026-09-07T14:32:03.712995Z  INFO modelled expanded operator plan finalist=0 modelled_cycles=364251 estimated_cycles=451577 estimated_exchange_cycles=124712
2026-09-07T14:32:04.352098Z  INFO screened tile mappings by exchange resource load candidates=94 baseline_cycles=42842 candidate_cycles=42842 baseline_pressure=35262 candidate_pressure=35262
2026-09-07T14:32:04.370944Z  INFO modelled expanded operator plan finalist=3 modelled_cycles=379076 estimated_cycles=468322 estimated_exchange_cycles=135688
2026-09-07T14:32:06.616248Z  INFO screened tile mappings by exchange resource load candidates=48 baseline_cycles=40842 candidate_cycles=40842 baseline_pressure=32638 candidate_pressure=32638
2026-09-07T14:32:06.626641Z  INFO modelled expanded operator plan finalist=2 modelled_cycles=214465 estimated_cycles=284367 estimated_exchange_cycles=114344
2026-09-07T14:32:11.243811Z  INFO screened tile mappings by exchange resource load candidates=38 baseline_cycles=44128 candidate_cycles=44128 baseline_pressure=37358 candidate_pressure=37358
2026-09-07T14:32:11.251779Z  INFO modelled expanded operator plan finalist=1 modelled_cycles=212419 estimated_cycles=295219 estimated_exchange_cycles=129928
2026-09-07T14:32:11.439898Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: prepared large physical exchange phase phase=1 transfers=5704 maximum_identical_destinations=1
2026-09-07T14:32:11.439961Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: optimized physical exchange schedule phase=1 initial_horizon=4454 selected_horizon=3492 endpoint_lower_bound=3360 lower_bound_gap=132 selected_kind="critical-neighborhood" neighborhood_improvements=1
2026-09-07T14:32:11.443669Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: finished large physical exchange phase phase=1 tile=522 row_words=38 horizon=3492
2026-09-07T14:32:13.006803Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: prepared large physical exchange phase phase=4 transfers=38637 maximum_identical_destinations=1
2026-09-07T14:32:13.006882Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: optimized physical exchange schedule phase=4 initial_horizon=16681 selected_horizon=16676 endpoint_lower_bound=12096 lower_bound_gap=4580 selected_kind="matching-waves" neighborhood_improvements=0
2026-09-07T14:32:13.040919Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: finished large physical exchange phase phase=4 tile=1057 row_words=175 horizon=16676
2026-09-07T14:32:14.944774Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: prepared large physical exchange phase phase=0 transfers=26496 maximum_identical_destinations=1
2026-09-07T14:32:14.944845Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: optimized physical exchange schedule phase=0 initial_horizon=2612 selected_horizon=2612 endpoint_lower_bound=1152 lower_bound_gap=1460 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-07T14:32:14.978028Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: finished large physical exchange phase phase=0 tile=741 row_words=507 horizon=2612
2026-09-07T14:32:16.393164Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: prepared large physical exchange phase phase=2 transfers=42282 maximum_identical_destinations=1
2026-09-07T14:32:16.393260Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: optimized physical exchange schedule phase=2 initial_horizon=6226 selected_horizon=6226 endpoint_lower_bound=2240 lower_bound_gap=3986 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-07T14:32:16.435976Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: finished large physical exchange phase phase=2 tile=443 row_words=320 horizon=6226
2026-09-07T14:32:16.468528Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: scheduled operator-plan finalist finalist=1 mapped=false analytical_cycles=295219 analytical_exchange_cycles=129928 scheduled_exchange_cycles=55173 refined_cycles=220464
2026-09-07T14:32:16.750181Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected physically scheduled operator-plan finalist selected=1 mapped=false
2026-09-07T14:32:16.750231Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: package build phase finished phase="select_finalist" elapsed_ms=16249 success=true
2026-09-07T14:32:16.750245Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=34722 exchange_phases=5
2026-09-07T14:32:16.750564Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="compile_runtime"}: package build phase finished phase="compile_runtime" elapsed_ms=0 success=true
2026-09-07T14:32:16.753449Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="plan_kernels"}: package build phase finished phase="plan_kernels" elapsed_ms=2 success=true
2026-09-07T14:32:16.753690Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="compile_kernels"}: compiling kernel source=/srv/home/gc-sdk/ipu-stack/crates/ipu-tests/../../device/gemm_f16_amp.S name="gemm_f8_interleaved_k160_c576_r26_r28" target=ipu21
2026-09-07T14:32:16.815957Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="compile_kernels"}: kernel artifact created object=/srv/home/.cache/ipu-stack/kernels/6ecfaef36cf1d2dbd283525a8b20079e3e8dfe740744e7b62a9b5b23f7cc6d31/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/6ecfaef36cf1d2dbd283525a8b20079e3e8dfe740744e7b62a9b5b23f7cc6d31/kernel.json
2026-09-07T14:32:16.817664Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="compile_kernels"}: package build phase finished phase="compile_kernels" elapsed_ms=64 success=true
2026-09-07T14:32:16.820977Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="link_runtime"}: package build phase finished phase="link_runtime" elapsed_ms=0 success=true
2026-09-07T14:32:17.759524Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="size_tile_code"}: package build phase finished phase="size_tile_code" elapsed_ms=17 success=true
2026-09-07T14:32:17.759624Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=318060 profile_bytes=312 exchange_table_bytes=3772 host_code_bytes=4392 generated_code_bytes=4100 code_address=322520 standard_ranges=[(409600, 524288)]
2026-09-07T14:32:17.808456Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="place_storage"}: package build phase finished phase="place_storage" elapsed_ms=48 success=true
2026-09-07T14:32:17.844068Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=3
2026-09-07T14:32:17.870933Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=1
2026-09-07T14:32:17.874240Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=1 transfers=5704 maximum_identical_destinations=1
2026-09-07T14:32:17.874267Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=1 initial_horizon=3492 selected_horizon=3492 endpoint_lower_bound=3360 lower_bound_gap=132 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:17.878090Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=1 tile=522 row_words=38 horizon=3492
2026-09-07T14:32:18.178765Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=4
2026-09-07T14:32:18.200843Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=4 transfers=38637 maximum_identical_destinations=1
2026-09-07T14:32:18.200885Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=4 initial_horizon=16676 selected_horizon=16676 endpoint_lower_bound=12096 lower_bound_gap=4580 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:18.230601Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=4 tile=1057 row_words=175 horizon=16676
2026-09-07T14:32:18.669925Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=0
2026-09-07T14:32:18.712765Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=0 transfers=26496 maximum_identical_destinations=1
2026-09-07T14:32:18.712810Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=0 initial_horizon=2612 selected_horizon=2612 endpoint_lower_bound=1152 lower_bound_gap=1460 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:18.743558Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=0 tile=741 row_words=507 horizon=2612
2026-09-07T14:32:20.325302Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=2
2026-09-07T14:32:20.350547Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=2 transfers=42282 maximum_identical_destinations=1
2026-09-07T14:32:20.350587Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=2 initial_horizon=6226 selected_horizon=6226 endpoint_lower_bound=2240 lower_bound_gap=3986 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:20.376174Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=2 tile=443 row_words=320 horizon=6226
2026-09-07T14:32:20.399562Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: package build phase finished phase="lower_exchanges" elapsed_ms=2591 success=true
2026-09-07T14:32:20.933270Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: reused exchange optimization after validating relocated rows phase=3
2026-09-07T14:32:20.968521Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: reused exchange optimization after validating relocated rows phase=1
2026-09-07T14:32:20.972150Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: prepared large physical exchange phase phase=1 transfers=5704 maximum_identical_destinations=1
2026-09-07T14:32:20.972179Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: optimized physical exchange schedule phase=1 initial_horizon=3492 selected_horizon=3492 endpoint_lower_bound=3360 lower_bound_gap=132 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:20.976270Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: finished large physical exchange phase phase=1 tile=522 row_words=38 horizon=3492
2026-09-07T14:32:21.287638Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: reused exchange optimization after validating relocated rows phase=4
2026-09-07T14:32:21.310854Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: prepared large physical exchange phase phase=4 transfers=38637 maximum_identical_destinations=1
2026-09-07T14:32:21.310885Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: optimized physical exchange schedule phase=4 initial_horizon=16676 selected_horizon=16676 endpoint_lower_bound=12096 lower_bound_gap=4580 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:21.338386Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: finished large physical exchange phase phase=4 tile=1057 row_words=175 horizon=16676
2026-09-07T14:32:23.577464Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: reused exchange optimization after validating relocated rows phase=2
2026-09-07T14:32:23.616198Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: prepared large physical exchange phase phase=2 transfers=42282 maximum_identical_destinations=1
2026-09-07T14:32:23.616273Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: optimized physical exchange schedule phase=2 initial_horizon=6226 selected_horizon=6226 endpoint_lower_bound=2240 lower_bound_gap=3986 selected_kind="reused" neighborhood_improvements=0
2026-09-07T14:32:23.639186Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: finished large physical exchange phase phase=2 tile=443 row_words=320 horizon=6226
2026-09-07T14:32:33.195034Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: prepared large physical exchange phase phase=0 transfers=26496 maximum_identical_destinations=1
2026-09-07T14:32:33.195120Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: optimized physical exchange schedule phase=0 initial_horizon=1323 selected_horizon=1323 endpoint_lower_bound=1152 lower_bound_gap=171 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-07T14:32:33.239808Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: finished large physical exchange phase phase=0 tile=741 row_words=576 horizon=1323
2026-09-07T14:32:33.299016Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: evaluated exchange placement finalist offset=24576 baseline_score=56395936768 score=53465691136 baseline_cycles=52173 candidate_cycles=50884 row_bytes=4056 row_capacity=16320 accepted=true
2026-09-07T14:32:33.344553Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: package build phase finished phase="optimize_exchange_placement" elapsed_ms=12944 success=true
2026-09-07T14:32:33.364477Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: costed final placed program final_cycles=219175 final_exchange=53884
2026-09-07T14:32:33.875229Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="prepare_tile_code"}: package build phase finished phase="prepare_tile_code" elapsed_ms=29 success=true
2026-09-07T14:32:33.910811Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="emit_tile_code"}: package build phase finished phase="emit_tile_code" elapsed_ms=35 success=true
2026-09-07T14:32:34.067803Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="build_tile_images"}: package build phase finished phase="build_tile_images" elapsed_ms=141 success=true
2026-09-07T14:32:34.179440Z  INFO writing application package tiles=1472
2026-09-07T14:32:34.280986Z  INFO application package written
2026-09-07T14:32:34.282730Z  INFO reading application package
2026-09-07T14:32:34.466005Z  INFO application package read tiles=1472 compiler=0.1.0
2026-09-07T14:32:34.469799Z  INFO waiting for hardware ownership path=artifacts/layout-sweep/device.lock
2026-09-07T14:32:34.469836Z  INFO acquired hardware ownership path=artifacts/layout-sweep/device.lock
2026-09-07T14:32:34.469844Z  INFO opening IPU device path="/dev/ipu0"
2026-09-07T14:32:34.469889Z  INFO IPU device attached path=/dev/ipu0
2026-09-07T14:32:34.469895Z  INFO initializing IPU device
2026-09-07T14:32:34.477754Z  INFO IPU device initialized
2026-09-07T14:32:34.477764Z  INFO replaying device configuration register_writes=2003
2026-09-07T14:32:34.489629Z  INFO loading application tile_count=1472 final_mark=23
2026-09-07T14:32:34.730241Z  INFO application loaded and initial synchronization completed tile_count=1472
2026-09-07T14:32:51.183509Z  INFO starting host exchange session startup_mark=23
2026-09-07T14:32:51.183661Z  INFO attaching host exchange pages pages=2
2026-09-07T14:32:51.184307Z  INFO host exchange session started
2026-09-07T14:32:51.185793Z  INFO invoking streaming host exchange call call="initialize" command=0 phases=122
2026-09-07T14:32:51.269294Z  INFO invoking streaming host exchange call call="run" command=0 phases=13
profile=artifacts/fp8/mlp-b1-final/profile.ipuprof tiles=1472 samples=31157
referenceMaximumAbsoluteError=0.025513 numericalTest=PASS
package=artifacts/fp8/mlp-b1-final/model.ipuexe tiles=1472 hardwareTest=PASS