2026-09-06T11:47:15.042050Z INFO applied GEMM plan constraint source_operation=0 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 2, column_partitions: 135, inner_partitions: 5, result_row_partitions: 5, result_column_partitions: 1, output_column_block: 32, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:47:15.553868Z 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: Ipu21Standard, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:47:16.098524Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=310833 estimated_exchange_cycles=119800 exchange_row_bytes=26300 peak_standard_bytes=234172 peak_interleaved_bytes=60416 peak_total_bytes=282300 maximum_standard_allocation_bytes=180224 active_tile_counts={1350, 1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:47:16.104870Z 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=1349 success=true 2026-09-06T11:47:19.003953Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=250218 estimated_exchange_cycles=66760 2026-09-06T11:47:19.086908Z 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 2026-09-06T11:47:19.088645Z 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=2983 success=true 2026-09-06T11:47:19.088671Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=25326 exchange_phases=4 2026-09-06T11:47:19.088854Z 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 2026-09-06T11:47:19.089406Z 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 2026-09-06T11:47:19.089722Z 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=0 success=true 2026-09-06T11:47:19.091581Z 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 2026-09-06T11:47:19.203101Z 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=111 success=true 2026-09-06T11:47:19.483056Z 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=12105 maximum_identical_destinations=1 2026-09-06T11:47:19.483086Z 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=6541 selected_horizon=5157 endpoint_lower_bound=4800 lower_bound_gap=357 selected_kind="matching-waves" neighborhood_improvements=0 2026-09-06T11:47:19.500286Z 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=599 row_words=66 horizon=5157 2026-09-06T11:47:24.502840Z 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=2887 ordinary_horizon=40644 paired_horizon=37871 use_paired=true 2026-09-06T11:47:24.565680Z 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=4217 maximum_identical_destinations=1 2026-09-06T11:47:24.565704Z 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=41723 selected_horizon=37871 endpoint_lower_bound=33120 lower_bound_gap=4751 selected_kind="critical-neighborhood" neighborhood_improvements=1 2026-09-06T11:47:24.596137Z 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=883 row_words=229 horizon=37871 2026-09-06T11:47:30.963096Z 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 2026-09-06T11:47:30.963146Z 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 2026-09-06T11:47:31.023968Z 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 2026-09-06T11:47:31.045541Z 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=11842 success=true 2026-09-06T11:47:31.890555Z 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 2026-09-06T11:47:31.890586Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319128 profile_bytes=132 exchange_table_bytes=2048 host_code_bytes=3476 generated_code_bytes=1744 code_address=322608 standard_ranges=[(409600, 524288)] 2026-09-06T11:47:31.992664Z 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=102 success=true 2026-09-06T11:47:31.997688Z 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 2026-09-06T11:47:32.075489Z 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 2026-09-06T11:47:32.080308Z 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=12105 maximum_identical_destinations=1 2026-09-06T11:47:32.080319Z 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=5157 selected_horizon=5157 endpoint_lower_bound=4800 lower_bound_gap=357 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:47:32.090718Z 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=599 row_words=66 horizon=5157 2026-09-06T11:47:32.696376Z 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 2026-09-06T11:47:32.714222Z 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=4217 maximum_identical_destinations=1 2026-09-06T11:47:32.714237Z 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=37871 selected_horizon=37871 endpoint_lower_bound=33120 lower_bound_gap=4751 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:47:32.743642Z 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=883 row_words=229 horizon=37871 2026-09-06T11:47:39.328876Z 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 2026-09-06T11:47:39.328925Z 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 2026-09-06T11:47:39.390049Z 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 2026-09-06T11:47:39.412008Z 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=7419 success=true 2026-09-06T11:47:40.679356Z 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 2026-09-06T11:47:40.790393Z 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 2026-09-06T11:47:40.795405Z 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=12105 maximum_identical_destinations=1 2026-09-06T11:47:40.795419Z 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=5157 selected_horizon=5157 endpoint_lower_bound=4800 lower_bound_gap=357 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:47:40.805583Z 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=599 row_words=66 horizon=5157 2026-09-06T11:47:47.041869Z 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=2887 ordinary_horizon=37433 paired_horizon=37465 use_paired=false 2026-09-06T11:47:47.097900Z 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=4217 maximum_identical_destinations=1 2026-09-06T11:47:47.097920Z 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=37433 selected_horizon=37433 endpoint_lower_bound=33408 lower_bound_gap=4025 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:47:47.126141Z 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=1030 row_words=197 horizon=37433 2026-09-06T11:47:53.080171Z 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 2026-09-06T11:47:53.080234Z 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 2026-09-06T11:47:53.134939Z 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 2026-09-06T11:47:53.171397Z 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=173497614336 score=130699395072 baseline_cycles=55073 candidate_cycles=54635 row_bytes=2056 row_capacity=16320 accepted=true 2026-09-06T11:47:53.187285Z 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=13775 success=true 2026-09-06T11:47:53.455142Z 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=11 success=true 2026-09-06T11:47:53.463948Z 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=8 success=true 2026-09-06T11:47:53.530628Z 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=63 success=true 2026-09-06T11:47:53.584426Z INFO writing application package tiles=1472 2026-09-06T11:47:53.619258Z INFO application package written 2026-09-06T11:47:53.620992Z INFO reading application package 2026-09-06T11:47:53.687718Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:47:53.688884Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:47:59.270353Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:47:59.270417Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:47:59.270489Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:47:59.270496Z INFO initializing IPU device 2026-09-06T11:47:59.278587Z INFO IPU device initialized 2026-09-06T11:47:59.278604Z INFO replaying device configuration register_writes=2003 2026-09-06T11:47:59.289529Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:47:59.606549Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:48:03.051584Z INFO starting host exchange session startup_mark=23 2026-09-06T11:48:03.051616Z INFO attaching host exchange pages pages=2 2026-09-06T11:48:03.052382Z INFO host exchange session started 2026-09-06T11:48:03.052471Z INFO invoking streaming host exchange call call="initialize" command=0 phases=14 2026-09-06T11:48:03.059260Z INFO invoking streaming host exchange call call="run" command=0 phases=93 profile=artifacts/layout-sweep/cross-cropped-2-0/execution.ipuprofile tiles=1472 samples=25154 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=236520000 weightBytes=20282880 cycles=213516 minimumTileCycles=190440 deviceMicroseconds=142.344 effectiveGemmTflops=101.572 peakTflops=282.624 efficiencyPercent=35.94 maximumAbsoluteError=0.011719 package=artifacts/layout-sweep/cross-cropped-2-0/model.ipuexe tiles=1472 hardwareTest=PASS