2026-09-06T11:22:57.816118Z INFO applied GEMM plan constraint source_operation=0 matching_plans=2 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 4, column_partitions: 45, inner_partitions: 8, result_row_partitions: 8, result_column_partitions: 1, output_column_block: 96, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:22:58.214663Z INFO applied GEMM plan constraint source_operation=2 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 2, orientation: Normal, row_partitions: 4, column_partitions: 24, inner_partitions: 15, result_row_partitions: 15, result_column_partitions: 1, output_column_block: 48, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:22:58.753047Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=342377 estimated_exchange_cycles=135800 exchange_row_bytes=29900 peak_standard_bytes=144588 peak_interleaved_bytes=120832 peak_total_bytes=216268 maximum_standard_allocation_bytes=114688 active_tile_counts={1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:22:58.758104Z 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=1232 success=true 2026-09-06T11:23:02.164978Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=281530 estimated_exchange_cycles=79944 2026-09-06T11:23:02.250238Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: eliminated finite GEMM padding clears using load-time SRAM initialization removed=72 2026-09-06T11:23:02.251995Z 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=3493 success=true 2026-09-06T11:23:02.252016Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=25821 exchange_phases=4 2026-09-06T11:23:02.252196Z 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 2026-09-06T11:23:02.252800Z 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=0 success=true 2026-09-06T11:23:02.252882Z 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_f16_interleaved_k144_c80_r182_r184" target=ipu21 2026-09-06T11:23:02.307534Z 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/b767e9e0f5b99536b23ec53f337a95ea35de862374ebfbd8d6e6d15e7d9cc445/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/b767e9e0f5b99536b23ec53f337a95ea35de862374ebfbd8d6e6d15e7d9cc445/kernel.json 2026-09-06T11:23:02.307752Z 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_f16_interleaved_k144_c96_r182_r184" target=ipu21 2026-09-06T11:23:02.357192Z 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/a44ea89325828806650044cd84aa030d9e7c657a56c9aaafb82bb4ffc45ec17e/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/a44ea89325828806650044cd84aa030d9e7c657a56c9aaafb82bb4ffc45ec17e/kernel.json 2026-09-06T11:23:02.357699Z 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=104 success=true 2026-09-06T11:23:02.360697Z 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 2026-09-06T11:23:02.484607Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="plan_exchange_storage"}: package build phase finished phase="plan_exchange_storage" elapsed_ms=123 success=true 2026-09-06T11:23:09.225015Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: prepared large physical exchange phase phase=1 transfers=66712 maximum_identical_destinations=1 2026-09-06T11:23:09.225062Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: optimized physical exchange schedule phase=1 initial_horizon=12347 selected_horizon=12347 endpoint_lower_bound=8448 lower_bound_gap=3899 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:23:09.293768Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: finished large physical exchange phase phase=1 tile=691 row_words=343 horizon=12347 2026-09-06T11:23:33.095928Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: compared ordinary and paired exchange schedules phase=2 candidates=7784 ordinary_horizon=57339 paired_horizon=55740 use_paired=true 2026-09-06T11:23:33.204956Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: prepared large physical exchange phase phase=2 transfers=10135 maximum_identical_destinations=1 2026-09-06T11:23:33.204979Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: optimized physical exchange schedule phase=2 initial_horizon=55740 selected_horizon=55740 endpoint_lower_bound=22464 lower_bound_gap=33276 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:23:33.267960Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: finished large physical exchange phase phase=2 tile=782 row_words=510 horizon=55740 2026-09-06T11:23:37.155551Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: prepared large physical exchange phase phase=3 transfers=63720 maximum_identical_destinations=1 2026-09-06T11:23:37.155584Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: optimized physical exchange schedule phase=3 initial_horizon=7991 selected_horizon=7991 endpoint_lower_bound=4704 lower_bound_gap=3287 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:23:37.193190Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: finished large physical exchange phase phase=3 tile=1058 row_words=294 horizon=7991 2026-09-06T11:23:37.208035Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: package build phase finished phase="lower_exchanges_provisional" elapsed_ms=34723 success=true 2026-09-06T11:23:37.446590Z 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=5 success=true 2026-09-06T11:23:37.446616Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319176 profile_bytes=132 exchange_table_bytes=4068 host_code_bytes=1580 generated_code_bytes=1468 code_address=320760 standard_ranges=[(409600, 524288)] 2026-09-06T11:23:37.555618Z 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=108 success=true 2026-09-06T11:23:37.563732Z 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 2026-09-06T11:23:44.821855Z 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=66712 maximum_identical_destinations=1 2026-09-06T11:23:44.821896Z 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=12347 selected_horizon=12347 endpoint_lower_bound=8448 lower_bound_gap=3899 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:23:44.890019Z 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=691 row_words=343 horizon=12347 2026-09-06T11:24:09.224751Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: compared ordinary and paired exchange schedules phase=2 candidates=7784 ordinary_horizon=57339 paired_horizon=55740 use_paired=true 2026-09-06T11:24:09.346692Z 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=10135 maximum_identical_destinations=1 2026-09-06T11:24:09.346718Z 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=55740 selected_horizon=55740 endpoint_lower_bound=22464 lower_bound_gap=33276 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:24:09.409483Z 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=782 row_words=510 horizon=55740 2026-09-06T11:24:15.680788Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=3 transfers=63720 maximum_identical_destinations=1 2026-09-06T11:24:15.680832Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=3 initial_horizon=7991 selected_horizon=7991 endpoint_lower_bound=4704 lower_bound_gap=3287 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:24:15.746636Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=3 tile=1058 row_words=294 horizon=7991 2026-09-06T11:24:15.766670Z 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=38211 success=true 2026-09-06T11:24:17.257892Z 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=0 2026-09-06T11:24:24.609689Z 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=66712 maximum_identical_destinations=1 2026-09-06T11:24:24.609734Z 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=12347 selected_horizon=12347 endpoint_lower_bound=8448 lower_bound_gap=3899 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:24:24.678635Z 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=691 row_words=343 horizon=12347 2026-09-06T11:24:47.926845Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: compared ordinary and paired exchange schedules phase=2 candidates=7784 ordinary_horizon=40797 paired_horizon=34816 use_paired=true 2026-09-06T11:24:48.059796Z 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=10135 maximum_identical_destinations=1 2026-09-06T11:24:48.059834Z 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=34816 selected_horizon=34816 endpoint_lower_bound=22464 lower_bound_gap=12352 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:24:48.127554Z 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=892 row_words=513 horizon=34816 2026-09-06T11:24:55.917103Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: prepared large physical exchange phase phase=3 transfers=63720 maximum_identical_destinations=1 2026-09-06T11:24:55.917169Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: optimized physical exchange schedule phase=3 initial_horizon=7991 selected_horizon=7991 endpoint_lower_bound=4704 lower_bound_gap=3287 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:24:55.986875Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: finished large physical exchange phase phase=3 tile=1058 row_words=294 horizon=7991 2026-09-06T11:24:56.030014Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: evaluated exchange placement finalist offset=28672 baseline_score=129230733312 score=115566870528 baseline_cycles=83122 candidate_cycles=62198 row_bytes=4012 row_capacity=16320 accepted=true 2026-09-06T11:24:56.058977Z 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=40292 success=true 2026-09-06T11:24:56.273375Z 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=13 success=true 2026-09-06T11:24:56.288297Z 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=14 success=true 2026-09-06T11:24:56.397722Z 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=102 success=true 2026-09-06T11:24:56.472848Z INFO writing application package tiles=1472 2026-09-06T11:24:56.549548Z INFO application package written 2026-09-06T11:24:56.551941Z INFO reading application package 2026-09-06T11:24:56.667475Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:24:56.670783Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:24:56.670839Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:24:56.670846Z INFO initializing IPU device 2026-09-06T11:24:56.679180Z INFO IPU device initialized 2026-09-06T11:24:56.679192Z INFO replaying device configuration register_writes=2003 2026-09-06T11:24:56.687121Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:24:56.992229Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:24:58.879104Z INFO starting host exchange session startup_mark=23 2026-09-06T11:24:58.879169Z INFO attaching host exchange pages pages=2 2026-09-06T11:24:58.879888Z INFO host exchange session started 2026-09-06T11:24:58.879959Z INFO invoking streaming host exchange call call="initialize" command=0 phases=14 2026-09-06T11:24:58.887734Z INFO invoking streaming host exchange call call="run" command=0 phases=33 profile=artifacts/layout-sweep/up-04/execution.ipuprofile tiles=1472 samples=25583 workload=siglip-mlp-f16-b1-t729-d1152-h4304-n1 benchmark=siglip-mlp-f16 batch=1 tokens=729 rows=729 dimension=1152 hiddenDimension=4304 blocks=1 biases=false activeTiles=1440 executionTiles=1472 inputBytes=75686400 weightBytes=19869696 cycles=207720 minimumTileCycles=198558 deviceMicroseconds=138.480 effectiveGemmTflops=104.406 peakTflops=282.624 efficiencyPercent=36.94 maximumAbsoluteError=0.009766 package=artifacts/layout-sweep/up-04/model.ipuexe tiles=1472 hardwareTest=PASS