2026-09-06T12:40:15.769728Z  INFO applied GEMM plan constraint source_operation=0 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 4, column_partitions: 92, inner_partitions: 4, result_row_partitions: 2, result_column_partitions: 2, output_column_block: 48, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct }
2026-09-06T12:40:16.443340Z  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: 5, result_column_partitions: 3, output_column_block: 48, weight_memory_class: Ipu21Standard, reduction_staging: Complete, local_weight_staging: Direct }
2026-09-06T12:40:16.995918Z  INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=265516 estimated_exchange_cycles=125560 exchange_row_bytes=27596 peak_standard_bytes=169932 peak_interleaved_bytes=60416 peak_total_bytes=230348 maximum_standard_allocation_bytes=114688 active_tile_counts={1440, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(2))]
2026-09-06T12:40:17.002073Z  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=1467 success=true
2026-09-06T12:40:20.789699Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=192446 estimated_exchange_cycles=67296
2026-09-06T12:40:20.936474Z  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
2026-09-06T12:40:20.941111Z  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=3938 success=true
2026-09-06T12:40:20.941144Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=28088 exchange_phases=4
2026-09-06T12:40:20.941441Z  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
2026-09-06T12:40:20.948705Z  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=7 success=true
2026-09-06T12:40:20.949084Z  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
2026-09-06T12:40:20.956101Z  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
2026-09-06T12:40:21.797428Z  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=841 success=true
2026-09-06T12:40:22.351456Z  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=10728 maximum_identical_destinations=1
2026-09-06T12:40:22.351497Z  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=8944 selected_horizon=8944 endpoint_lower_bound=5888 lower_bound_gap=3056 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:22.362478Z  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=966 row_words=67 horizon=8944
2026-09-06T12:40:25.613602Z  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=2041 ordinary_horizon=39938 paired_horizon=28096 use_paired=true
2026-09-06T12:40:25.665067Z  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=3410 maximum_identical_destinations=1
2026-09-06T12:40:25.665113Z  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=28096 selected_horizon=28096 endpoint_lower_bound=26784 lower_bound_gap=1312 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:25.690280Z  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=934 row_words=164 horizon=28096
2026-09-06T12:40:27.328396Z  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=24760 maximum_identical_destinations=1
2026-09-06T12:40:27.328448Z  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=5941 selected_horizon=5941 endpoint_lower_bound=4560 lower_bound_gap=1381 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:27.356265Z  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=718 row_words=132 horizon=5941
2026-09-06T12:40:27.368486Z  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=5571 success=true
2026-09-06T12:40:28.114342Z  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=6 success=true
2026-09-06T12:40:28.114390Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=317036 profile_bytes=128 exchange_table_bytes=1276 host_code_bytes=2400 generated_code_bytes=1400 code_address=319440 standard_ranges=[(409600, 524288)]
2026-09-06T12:40:28.265567Z  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=151 success=true
2026-09-06T12:40:28.278552Z  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
2026-09-06T12:40:28.774467Z  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=10728 maximum_identical_destinations=1
2026-09-06T12:40:28.774513Z  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=8944 selected_horizon=8944 endpoint_lower_bound=5888 lower_bound_gap=3056 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:28.790800Z  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=966 row_words=67 horizon=8944
2026-09-06T12:40:32.712482Z  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=2041 ordinary_horizon=39938 paired_horizon=28096 use_paired=true
2026-09-06T12:40:32.764627Z  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=3410 maximum_identical_destinations=1
2026-09-06T12:40:32.764661Z  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=28096 selected_horizon=28096 endpoint_lower_bound=26784 lower_bound_gap=1312 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:32.789165Z  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=934 row_words=164 horizon=28096
2026-09-06T12:40:35.076485Z  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=24760 maximum_identical_destinations=1
2026-09-06T12:40:35.076529Z  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=5941 selected_horizon=5941 endpoint_lower_bound=4560 lower_bound_gap=1381 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:35.118378Z  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=718 row_words=132 horizon=5941
2026-09-06T12:40:35.133263Z  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=6867 success=true
2026-09-06T12:40:37.650244Z  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
2026-09-06T12:40:37.778802Z  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
2026-09-06T12:40:37.784846Z  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=10728 maximum_identical_destinations=1
2026-09-06T12:40:37.784865Z  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=8944 selected_horizon=8944 endpoint_lower_bound=5888 lower_bound_gap=3056 selected_kind="reused" neighborhood_improvements=0
2026-09-06T12:40:37.795791Z  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=966 row_words=67 horizon=8944
2026-09-06T12:40:41.363241Z  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=2041 ordinary_horizon=37412 paired_horizon=28065 use_paired=true
2026-09-06T12:40:41.411605Z  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=3410 maximum_identical_destinations=1
2026-09-06T12:40:41.411639Z  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=28065 selected_horizon=28065 endpoint_lower_bound=26784 lower_bound_gap=1281 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:41.447806Z  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=933 row_words=167 horizon=28065
2026-09-06T12:40:43.084897Z  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=24760 maximum_identical_destinations=1
2026-09-06T12:40:43.084945Z  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=5941 selected_horizon=5941 endpoint_lower_bound=4560 lower_bound_gap=1381 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T12:40:43.112881Z  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=718 row_words=132 horizon=5941
2026-09-06T12:40:43.135198Z  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=170977886208 score=150597992448 baseline_cycles=50029 candidate_cycles=49998 row_bytes=1272 row_capacity=16320 accepted=true
2026-09-06T12:40:43.152656Z  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=8019 success=true
2026-09-06T12:40:43.659850Z  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=109 success=true
2026-09-06T12:40:43.674793Z  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
2026-09-06T12:40:43.824314Z  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
2026-09-06T12:40:43.933507Z  INFO writing application package tiles=1472
2026-09-06T12:40:43.984547Z  INFO application package written
2026-09-06T12:40:43.985054Z  INFO reading application package
2026-09-06T12:40:44.087735Z  INFO application package read tiles=1472 compiler=0.1.0
2026-09-06T12:40:44.088084Z  INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/alternatives-followup/device.lock
2026-09-06T12:40:50.545923Z  INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/alternatives-followup/device.lock
2026-09-06T12:40:50.545968Z  INFO opening IPU device path="/dev/ipu0"
2026-09-06T12:40:50.546023Z  INFO IPU device attached path=/dev/ipu0
2026-09-06T12:40:50.546027Z  INFO initializing IPU device
2026-09-06T12:40:50.558002Z  INFO IPU device initialized
2026-09-06T12:40:50.558033Z  INFO replaying device configuration register_writes=2003
2026-09-06T12:40:50.629882Z  INFO loading application tile_count=1472 final_mark=23
2026-09-06T12:40:50.980016Z  INFO application loaded and initial synchronization completed tile_count=1472
2026-09-06T12:40:54.168856Z  INFO starting host exchange session startup_mark=23
2026-09-06T12:40:54.168906Z  INFO attaching host exchange pages pages=2
2026-09-06T12:40:54.169402Z  INFO host exchange session started
2026-09-06T12:40:54.169513Z  INFO invoking streaming host exchange call call="initialize" command=0 phases=14
2026-09-06T12:40:54.178454Z  INFO invoking streaming host exchange call call="run" command=0 phases=59
profile=artifacts/layout-sweep/alternatives-followup/both-mixed/execution.ipuprofile tiles=1472 samples=23302
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=154736640 weightBytes=19869696 cycles=184368 minimumTileCycles=173616 deviceMicroseconds=122.912 effectiveGemmTflops=117.630 peakTflops=282.624 efficiencyPercent=41.62 maximumAbsoluteError=0.011719
package=artifacts/layout-sweep/alternatives-followup/both-mixed/model.ipuexe tiles=1472 hardwareTest=PASS