2026-09-06T11:37:19.687639Z INFO applied GEMM plan constraint source_operation=0 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 3, column_partitions: 90, inner_partitions: 5, result_row_partitions: 5, result_column_partitions: 1, output_column_block: 48, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:37:20.113613Z 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 } 2026-09-06T11:37:20.653159Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=320918 estimated_exchange_cycles=124760 exchange_row_bytes=27416 peak_standard_bytes=158488 peak_interleaved_bytes=83456 peak_total_bytes=241944 maximum_standard_allocation_bytes=131072 active_tile_counts={1350, 1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:37:20.659070Z 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=1263 success=true 2026-09-06T11:37:23.923598Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=259265 estimated_exchange_cycles=70728 2026-09-06T11:37:24.007433Z 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 2026-09-06T11:37:24.009134Z 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=3350 success=true 2026-09-06T11:37:24.009150Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=25236 exchange_phases=4 2026-09-06T11:37:24.009349Z 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 2026-09-06T11:37:24.009911Z 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 2026-09-06T11:37:24.010047Z 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_k240_c32_r242_r244" target=ipu21 2026-09-06T11:37:24.061907Z 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/fe9a9df432e12cb077b1e12da3767010e4975053bad851456c8db993e07e06d9/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/fe9a9df432e12cb077b1e12da3767010e4975053bad851456c8db993e07e06d9/kernel.json 2026-09-06T11:37:24.062291Z 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_k240_c48_r242_r244" target=ipu21 2026-09-06T11:37:24.106072Z 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/d7d0992b3f171a3a14f839461aef2a98f8c0224258b92a40fcd6da28f069b8bb/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/d7d0992b3f171a3a14f839461aef2a98f8c0224258b92a40fcd6da28f069b8bb/kernel.json 2026-09-06T11:37:24.106635Z 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=96 success=true 2026-09-06T11:37:24.109904Z 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 2026-09-06T11:37:24.225039Z 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=115 success=true 2026-09-06T11:37:24.845268Z 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=18830 maximum_identical_destinations=1 2026-09-06T11:37:24.845302Z 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=6696 selected_horizon=6688 endpoint_lower_bound=4944 lower_bound_gap=1744 selected_kind="critical-neighborhood" neighborhood_improvements=1 2026-09-06T11:37:24.860613Z 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=390 row_words=96 horizon=6688 2026-09-06T11:37:29.741075Z 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=4179 ordinary_horizon=57836 paired_horizon=49384 use_paired=true 2026-09-06T11:37:29.817730Z 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=5562 maximum_identical_destinations=1 2026-09-06T11:37:29.817752Z 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=49384 selected_horizon=49384 endpoint_lower_bound=24192 lower_bound_gap=25192 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:37:29.858667Z 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=684 row_words=277 horizon=49384 2026-09-06T11:37:36.020304Z 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 2026-09-06T11:37:36.020347Z 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 2026-09-06T11:37:36.081640Z 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 2026-09-06T11:37:36.105856Z 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=11880 success=true 2026-09-06T11:37:36.780092Z 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 2026-09-06T11:37:36.780138Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319176 profile_bytes=132 exchange_table_bytes=2296 host_code_bytes=2580 generated_code_bytes=1560 code_address=321760 standard_ranges=[(409600, 524288)] 2026-09-06T11:37:36.882861Z 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 2026-09-06T11:37:36.892238Z 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 2026-09-06T11:37:37.010920Z 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 2026-09-06T11:37:37.018363Z 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=18830 maximum_identical_destinations=1 2026-09-06T11:37:37.018375Z 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=6688 selected_horizon=6688 endpoint_lower_bound=4944 lower_bound_gap=1744 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:37:37.034431Z 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=390 row_words=96 horizon=6688 2026-09-06T11:37:37.808356Z 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 2026-09-06T11:37:37.833948Z 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=5562 maximum_identical_destinations=1 2026-09-06T11:37:37.833964Z 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=49384 selected_horizon=49384 endpoint_lower_bound=24192 lower_bound_gap=25192 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:37:37.871250Z 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=684 row_words=277 horizon=49384 2026-09-06T11:37:44.514526Z 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 2026-09-06T11:37:44.514562Z 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 2026-09-06T11:37:44.575332Z 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 2026-09-06T11:37:44.595648Z 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=7712 success=true 2026-09-06T11:37:45.860937Z 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 2026-09-06T11:37:45.986054Z 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 2026-09-06T11:37:45.993754Z 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=18830 maximum_identical_destinations=1 2026-09-06T11:37:45.993779Z 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=6688 selected_horizon=6688 endpoint_lower_bound=4944 lower_bound_gap=1744 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:37:46.008231Z 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=390 row_words=96 horizon=6688 2026-09-06T11:37:51.704360Z 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=4179 ordinary_horizon=44822 paired_horizon=68413 use_paired=false 2026-09-06T11:37:51.758753Z 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=5562 maximum_identical_destinations=1 2026-09-06T11:37:51.758772Z 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=76970 selected_horizon=44822 endpoint_lower_bound=33408 lower_bound_gap=11414 selected_kind="critical-neighborhood" neighborhood_improvements=1 2026-09-06T11:37:51.781989Z 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=684 row_words=251 horizon=44822 2026-09-06T11:37:56.009035Z 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 2026-09-06T11:37:56.009064Z 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 2026-09-06T11:37:56.045671Z 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 2026-09-06T11:37:56.066284Z 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=448685899776 score=171905089536 baseline_cycles=69952 candidate_cycles=65390 row_bytes=2288 row_capacity=16320 accepted=true 2026-09-06T11:37:56.079129Z 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=11483 success=true 2026-09-06T11:37:56.269378Z 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=10 success=true 2026-09-06T11:37:56.280636Z 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=11 success=true 2026-09-06T11:37:56.387046Z 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 2026-09-06T11:37:56.455671Z INFO writing application package tiles=1472 2026-09-06T11:37:56.528875Z INFO application package written 2026-09-06T11:37:56.531139Z INFO reading application package 2026-09-06T11:37:56.628637Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:37:56.630718Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:37:56.630729Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:37:56.630733Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:37:56.630758Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:37:56.630760Z INFO initializing IPU device 2026-09-06T11:37:56.638594Z INFO IPU device initialized 2026-09-06T11:37:56.638608Z INFO replaying device configuration register_writes=2003 2026-09-06T11:37:56.645488Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:37:56.960987Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:37:59.039882Z INFO starting host exchange session startup_mark=23 2026-09-06T11:37:59.039913Z INFO attaching host exchange pages pages=2 2026-09-06T11:37:59.040640Z INFO host exchange session started 2026-09-06T11:37:59.040697Z INFO invoking streaming host exchange call call="initialize" command=0 phases=14 2026-09-06T11:37:59.046697Z INFO invoking streaming host exchange call call="run" command=0 phases=65 profile=artifacts/layout-sweep/up-12/execution.ipuprofile tiles=1472 samples=25220 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=157680000 weightBytes=20282880 cycles=219990 minimumTileCycles=203616 deviceMicroseconds=146.660 effectiveGemmTflops=98.583 peakTflops=282.624 efficiencyPercent=34.88 maximumAbsoluteError=0.009766 package=artifacts/layout-sweep/up-12/model.ipuexe tiles=1472 hardwareTest=PASS