2026-09-06T11:38:04.102364Z INFO applied GEMM plan constraint source_operation=0 matching_plans=2 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 10, column_partitions: 34, inner_partitions: 4, result_row_partitions: 4, result_column_partitions: 1, output_column_block: 128, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:38:04.485425Z 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 } 2026-09-06T11:38:05.027123Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=376583 estimated_exchange_cycles=154040 exchange_row_bytes=34004 peak_standard_bytes=148692 peak_interleaved_bytes=175104 peak_total_bytes=258260 maximum_standard_allocation_bytes=114688 active_tile_counts={1360, 1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:38:05.031216Z 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=1218 success=true 2026-09-06T11:38:11.378612Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=310778 estimated_exchange_cycles=96112 2026-09-06T11:38:11.466250Z 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 2026-09-06T11:38:11.468077Z 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=6436 success=true 2026-09-06T11:38:11.468095Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=24880 exchange_phases=4 2026-09-06T11:38:11.468306Z 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 2026-09-06T11:38:11.468893Z 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 2026-09-06T11:38:11.469106Z 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_k288_c112_r72_r74" target=ipu21 2026-09-06T11:38:11.519266Z 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/5d330e61c0027d2de5cb801c226ab2083c49eead2307a95cf44b069e12bf9d93/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/5d330e61c0027d2de5cb801c226ab2083c49eead2307a95cf44b069e12bf9d93/kernel.json 2026-09-06T11:38:11.519595Z 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_k288_c128_r72_r74" target=ipu21 2026-09-06T11:38:11.569388Z 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/02781c9f13679fc1048a1ab676e1d592b070ae1de4f2e7dc4f472bb6b7f49ce6/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/02781c9f13679fc1048a1ab676e1d592b070ae1de4f2e7dc4f472bb6b7f49ce6/kernel.json 2026-09-06T11:38:11.569733Z 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=100 success=true 2026-09-06T11:38:11.573575Z 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 2026-09-06T11:38:11.689888Z 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=116 success=true 2026-09-06T11:38:12.649601Z 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=36584 maximum_identical_destinations=1 2026-09-06T11:38:12.649634Z 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=4947 selected_horizon=4947 endpoint_lower_bound=4224 lower_bound_gap=723 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:38:12.675611Z 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=393 row_words=165 horizon=4947 2026-09-06T11:38:47.768468Z 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=9578 ordinary_horizon=45780 paired_horizon=40319 use_paired=true 2026-09-06T11:38:48.015878Z 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=12287 maximum_identical_destinations=1 2026-09-06T11:38:48.015902Z 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=40319 selected_horizon=40319 endpoint_lower_bound=22752 lower_bound_gap=17567 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:38:48.157890Z 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=574 row_words=694 horizon=40319 2026-09-06T11:38:54.292242Z 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 2026-09-06T11:38:54.292282Z 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 2026-09-06T11:38:54.359083Z 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 2026-09-06T11:38:54.382790Z 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=42692 success=true 2026-09-06T11:38:54.725261Z 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 2026-09-06T11:38:54.725295Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319128 profile_bytes=132 exchange_table_bytes=3936 host_code_bytes=2156 generated_code_bytes=1648 code_address=321288 standard_ranges=[(409600, 524288)] 2026-09-06T11:38:54.823358Z 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=98 success=true 2026-09-06T11:38:54.839513Z 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 2026-09-06T11:38:55.091538Z 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 2026-09-06T11:38:55.106065Z 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=36584 maximum_identical_destinations=1 2026-09-06T11:38:55.106077Z 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=4947 selected_horizon=4947 endpoint_lower_bound=4224 lower_bound_gap=723 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:38:55.133407Z 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=393 row_words=165 horizon=4947 2026-09-06T11:39:32.857427Z 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=9578 ordinary_horizon=45780 paired_horizon=40319 use_paired=true 2026-09-06T11:39:33.006655Z 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=12287 maximum_identical_destinations=1 2026-09-06T11:39:33.006674Z 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=40319 selected_horizon=40319 endpoint_lower_bound=22752 lower_bound_gap=17567 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:39:33.093714Z 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=574 row_words=694 horizon=40319 2026-09-06T11:39:37.443860Z 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 2026-09-06T11:39:37.443889Z 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 2026-09-06T11:39:37.486524Z 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 2026-09-06T11:39:37.502053Z 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=42678 success=true 2026-09-06T11:39:38.736959Z 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 2026-09-06T11:39:38.991937Z 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 2026-09-06T11:39:39.006765Z 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=36584 maximum_identical_destinations=1 2026-09-06T11:39:39.006781Z 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=4947 selected_horizon=4947 endpoint_lower_bound=4224 lower_bound_gap=723 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:39:39.034407Z 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=393 row_words=165 horizon=4947 2026-09-06T11:40:14.965990Z 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=9578 ordinary_horizon=40460 paired_horizon=36348 use_paired=true 2026-09-06T11:40:15.117603Z 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=12287 maximum_identical_destinations=1 2026-09-06T11:40:15.117622Z 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=36348 selected_horizon=36348 endpoint_lower_bound=22752 lower_bound_gap=13596 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:40:15.203761Z 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=574 row_words=692 horizon=36348 2026-09-06T11:40:19.547568Z 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 2026-09-06T11:40:19.547594Z 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 2026-09-06T11:40:19.588520Z 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 2026-09-06T11:40:19.613540Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: evaluated exchange placement finalist offset=4096 baseline_score=420600840192 score=172595183616 baseline_cycles=71812 candidate_cycles=67841 row_bytes=3944 row_capacity=16320 accepted=true 2026-09-06T11:40:19.624890Z 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=42122 success=true 2026-09-06T11:40:19.739304Z 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=6 success=true 2026-09-06T11:40:19.748484Z 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=9 success=true 2026-09-06T11:40:19.816314Z 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 2026-09-06T11:40:19.855387Z INFO writing application package tiles=1472 2026-09-06T11:40:19.903766Z INFO application package written 2026-09-06T11:40:19.905471Z INFO reading application package 2026-09-06T11:40:19.973984Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:40:19.975559Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:40:19.975568Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:40:19.975572Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:40:19.975595Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:40:19.975597Z INFO initializing IPU device 2026-09-06T11:40:19.983404Z INFO IPU device initialized 2026-09-06T11:40:19.983407Z INFO replaying device configuration register_writes=2003 2026-09-06T11:40:19.988932Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:40:20.223016Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:40:21.192504Z INFO starting host exchange session startup_mark=23 2026-09-06T11:40:21.192524Z INFO attaching host exchange pages pages=2 2026-09-06T11:40:21.193175Z INFO host exchange session started 2026-09-06T11:40:21.193216Z INFO invoking streaming host exchange call call="initialize" command=0 phases=36 2026-09-06T11:40:21.201921Z INFO invoking streaming host exchange call call="run" command=0 phases=29 profile=artifacts/layout-sweep/up-16/execution.ipuprofile tiles=1472 samples=25201 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=57185280 weightBytes=19869696 cycles=233064 minimumTileCycles=218046 deviceMicroseconds=155.376 effectiveGemmTflops=93.053 peakTflops=282.624 efficiencyPercent=32.92 maximumAbsoluteError=0.011719 package=artifacts/layout-sweep/up-16/model.ipuexe tiles=1472 hardwareTest=PASS