2026-09-06T16:41:23.325608Z INFO retained operator-plan finalist finalist=0 values=11 operations=7 estimated_cycles=569331 estimated_exchange_cycles=173640 exchange_row_bytes=35532 peak_standard_bytes=96716 peak_interleaved_bytes=114688 peak_total_bytes=166604 maximum_standard_allocation_bytes=16384 active_tile_counts={12, 64} padding_group_counts={} conversion_sources=[] 2026-09-06T16:41:23.325751Z INFO retained operator-plan finalist finalist=1 values=11 operations=7 estimated_cycles=579969 estimated_exchange_cycles=191640 exchange_row_bytes=38796 peak_standard_bytes=99980 peak_interleaved_bytes=44480 peak_total_bytes=132748 maximum_standard_allocation_bytes=16384 active_tile_counts={12, 30, 64} padding_group_counts={} conversion_sources=[] 2026-09-06T16:41:23.325774Z INFO retained operator-plan finalist finalist=2 values=11 operations=7 estimated_cycles=582050 estimated_exchange_cycles=187440 exchange_row_bytes=38244 peak_standard_bytes=99428 peak_interleaved_bytes=46592 peak_total_bytes=132196 maximum_standard_allocation_bytes=16384 active_tile_counts={12, 32, 64} padding_group_counts={} conversion_sources=[] 2026-09-06T16:41:23.325839Z INFO retained operator-plan finalist finalist=3 values=12 operations=8 estimated_cycles=585611 estimated_exchange_cycles=188320 exchange_row_bytes=38180 peak_standard_bytes=99364 peak_interleaved_bytes=73344 peak_total_bytes=139620 maximum_standard_allocation_bytes=16384 active_tile_counts={12, 30, 64} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T16:41:23.332678Z INFO ipu_codegen.package.phase{phase="lower_mid"}: package build phase finished phase="lower_mid" elapsed_ms=468 success=true 2026-09-06T16:41:23.796569Z INFO screened tile mappings by exchange resource load candidates=15 baseline_cycles=63344 candidate_cycles=63344 baseline_pressure=61298 candidate_pressure=61298 2026-09-06T16:41:23.800017Z INFO modelled expanded operator plan finalist=0 modelled_cycles=445149 estimated_cycles=466885 estimated_exchange_cycles=94080 2026-09-06T16:41:23.815990Z INFO screened tile mappings by exchange resource load candidates=19 baseline_cycles=68656 candidate_cycles=68656 baseline_pressure=64224 candidate_pressure=64224 2026-09-06T16:41:23.819864Z INFO modelled expanded operator plan finalist=2 modelled_cycles=506193 estimated_cycles=567449 estimated_exchange_cycles=140712 2026-09-06T16:41:23.827627Z INFO screened tile mappings by exchange resource load candidates=23 baseline_cycles=71376 candidate_cycles=71376 baseline_pressure=67290 candidate_pressure=67290 2026-09-06T16:41:23.828761Z INFO screened tile mappings by exchange resource load candidates=23 baseline_cycles=70952 candidate_cycles=70952 baseline_pressure=66149 candidate_pressure=66149 2026-09-06T16:41:23.831030Z INFO modelled expanded operator plan finalist=1 modelled_cycles=470163 estimated_cycles=531227 estimated_exchange_cycles=145040 2026-09-06T16:41:23.832456Z INFO modelled expanded operator plan finalist=3 modelled_cycles=459022 estimated_cycles=519270 estimated_exchange_cycles=143200 2026-09-06T16:41:23.919975Z INFO ipu_codegen.package.phase{phase="select_finalist"}: prepared large physical exchange phase phase=1 transfers=1038 maximum_identical_destinations=1 2026-09-06T16:41:23.920015Z INFO ipu_codegen.package.phase{phase="select_finalist"}: optimized physical exchange schedule phase=1 initial_horizon=1121 selected_horizon=1121 endpoint_lower_bound=736 lower_bound_gap=385 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T16:41:23.920978Z INFO ipu_codegen.package.phase{phase="select_finalist"}: finished large physical exchange phase phase=1 tile=49 row_words=162 horizon=1121 2026-09-06T16:41:24.076970Z INFO ipu_codegen.package.phase{phase="select_finalist"}: scheduled operator-plan finalist finalist=0 mapped=false analytical_cycles=466885 analytical_exchange_cycles=94080 scheduled_exchange_cycles=74373 refined_cycles=447178 2026-09-06T16:41:24.091725Z INFO ipu_codegen.package.phase{phase="select_finalist"}: selected physically scheduled operator-plan finalist selected=0 mapped=false 2026-09-06T16:41:24.091776Z INFO ipu_codegen.package.phase{phase="select_finalist"}: package build phase finished phase="select_finalist" elapsed_ms=759 success=true 2026-09-06T16:41:24.091789Z INFO lowered graph for package construction logical_shards=5300 exchange_phases=15 2026-09-06T16:41:24.092041Z INFO ipu_codegen.package.phase{phase="compile_runtime"}: package build phase finished phase="compile_runtime" elapsed_ms=0 success=true 2026-09-06T16:41:24.092711Z INFO ipu_codegen.package.phase{phase="plan_kernels"}: package build phase finished phase="plan_kernels" elapsed_ms=0 success=true 2026-09-06T16:41:24.093288Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: compiling kernel source=/srv/home/gc-sdk/ipu-stack/crates/ipu-tests/../../device/attention_stages_f16.S name="attention_softmax_d72_p64_tail" target=ipu21 2026-09-06T16:41:24.149922Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: kernel artifact created object=/srv/home/.cache/ipu-stack/kernels/f8d9f2e9ee14e37469a232e9e28949c4edba6c2f4974fd274b9678f6fd79afdf/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/f8d9f2e9ee14e37469a232e9e28949c4edba6c2f4974fd274b9678f6fd79afdf/kernel.json 2026-09-06T16:41:24.150351Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: compiling kernel source=/srv/home/gc-sdk/ipu-stack/crates/ipu-tests/../../device/attention_softmax_f16.S name="attention_softmax_d72_p64_full" target=ipu21 2026-09-06T16:41:24.203182Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: kernel artifact created object=/srv/home/.cache/ipu-stack/kernels/da0777b96831d9d99504f730aba0b33640bb55febb3f3cb54c4724907a9ebdd2/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/da0777b96831d9d99504f730aba0b33640bb55febb3f3cb54c4724907a9ebdd2/kernel.json 2026-09-06T16:41:24.203647Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: compiling kernel source=/srv/home/gc-sdk/ipu-stack/crates/ipu-tests/../../device/attention_stages_f16.S name="attention_merge_v72_p80_k64" target=ipu21 2026-09-06T16:41:24.253007Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: kernel artifact created object=/srv/home/.cache/ipu-stack/kernels/96393dee1fbf324c52d1bfc57090e17eb0469c69d00eead51bf19ba432220ddc/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/96393dee1fbf324c52d1bfc57090e17eb0469c69d00eead51bf19ba432220ddc/kernel.json 2026-09-06T16:41:24.253098Z INFO ipu_codegen.package.phase{phase="compile_kernels"}: package build phase finished phase="compile_kernels" elapsed_ms=160 success=true 2026-09-06T16:41:24.253623Z INFO ipu_codegen.package.phase{phase="link_runtime"}: package build phase finished phase="link_runtime" elapsed_ms=0 success=true 2026-09-06T16:41:24.277399Z INFO ipu_codegen.package.phase{phase="size_tile_code"}: package build phase finished phase="size_tile_code" elapsed_ms=4 success=true 2026-09-06T16:41:24.277436Z INFO allocated package support memory linked_end=322204 profile_bytes=0 exchange_table_bytes=3852 host_code_bytes=1076 generated_code_bytes=2148 code_address=323288 standard_ranges=[(393216, 524288)] 2026-09-06T16:41:24.284729Z INFO ipu_codegen.package.phase{phase="place_storage"}: package build phase finished phase="place_storage" elapsed_ms=7 success=true 2026-09-06T16:41:24.291727Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=0 2026-09-06T16:41:24.323584Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=1 2026-09-06T16:41:24.324103Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=1 transfers=1038 maximum_identical_destinations=1 2026-09-06T16:41:24.324115Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=1 initial_horizon=1121 selected_horizon=1121 endpoint_lower_bound=736 lower_bound_gap=385 selected_kind="reused" neighborhood_improvements=0 2026-09-06T16:41:24.325117Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=1 tile=49 row_words=162 horizon=1121 2026-09-06T16:41:24.348855Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=2 2026-09-06T16:41:24.369528Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=4 2026-09-06T16:41:24.375191Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=5 2026-09-06T16:41:24.380609Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=6 2026-09-06T16:41:24.386544Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=7 2026-09-06T16:41:24.392099Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=8 2026-09-06T16:41:24.398029Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=9 2026-09-06T16:41:24.403438Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=10 2026-09-06T16:41:24.408841Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=11 2026-09-06T16:41:24.414605Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=12 2026-09-06T16:41:24.419980Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: reused exchange optimization after validating relocated rows phase=13 2026-09-06T16:41:24.433859Z INFO ipu_codegen.package.phase{phase="lower_exchanges"}: package build phase finished phase="lower_exchanges" elapsed_ms=149 success=true 2026-09-06T16:41:24.519857Z INFO ipu_codegen.package.phase{phase="optimize_exchange_placement"}: package build phase finished phase="optimize_exchange_placement" elapsed_ms=85 success=true 2026-09-06T16:41:24.520619Z INFO costed final placed program final_cycles=447367 final_exchange=74562 2026-09-06T16:41:24.535424Z INFO ipu_codegen.package.phase{phase="prepare_tile_code"}: package build phase finished phase="prepare_tile_code" elapsed_ms=5 success=true 2026-09-06T16:41:24.542114Z INFO ipu_codegen.package.phase{phase="emit_tile_code"}: package build phase finished phase="emit_tile_code" elapsed_ms=6 success=true 2026-09-06T16:41:24.669733Z INFO ipu_codegen.package.phase{phase="build_tile_images"}: package build phase finished phase="build_tile_images" elapsed_ms=124 success=true 2026-09-06T16:41:24.678795Z INFO writing application package tiles=1472 2026-09-06T16:41:24.697589Z INFO application package written 2026-09-06T16:41:24.698181Z INFO reading application package 2026-09-06T16:41:24.720700Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T16:41:24.721645Z INFO waiting for hardware ownership path=artifacts/layout-sweep/device.lock 2026-09-06T16:41:24.721659Z INFO acquired hardware ownership path=artifacts/layout-sweep/device.lock 2026-09-06T16:41:24.721664Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T16:41:24.721697Z INFO IPU device attached path=/dev/ipu0 2026-09-06T16:41:24.721700Z INFO initializing IPU device 2026-09-06T16:41:24.729704Z INFO IPU device initialized 2026-09-06T16:41:24.729710Z INFO replaying device configuration register_writes=2003 2026-09-06T16:41:24.733656Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T16:41:24.921229Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T16:41:25.120065Z INFO starting host exchange session startup_mark=23 2026-09-06T16:41:25.120096Z INFO attaching host exchange pages pages=2 2026-09-06T16:41:25.120912Z INFO host exchange session started 2026-09-06T16:41:25.120924Z INFO invoking streaming host exchange call call="initialize" command=0 phases=24 2026-09-06T16:41:25.125067Z INFO invoking streaming host exchange call call="run" command=0 phases=5 checkpoint=0 operation=0 tensors=1 samples=256 maximumAbsoluteError=0.000244 checkpoint=1 operation=1 tensors=1 samples=256 maximumAbsoluteError=0.000244 checkpoint=2 operation=2 tensors=1 samples=256 maximumAbsoluteError=0.000244 checkpoint=3 operation=6 tensors=1 samples=256 maximumAbsoluteError=0.000011 diagnosticCheckpoints=4 sampleLimit=256 numericalTest=PASS package=artifacts/layout-sweep/attention-row-diagnostic/model.ipuexe tiles=1472 hardwareTest=PASS