2026-09-06T11:18:36.743987Z INFO applied GEMM plan constraint source_operation=0 matching_plans=2 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 7, column_partitions: 68, inner_partitions: 3, result_row_partitions: 3, result_column_partitions: 1, output_column_block: 64, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:18:37.162797Z 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 } 2026-09-06T11:18:37.707148Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=331895 estimated_exchange_cycles=134840 exchange_row_bytes=29684 peak_standard_bytes=144372 peak_interleaved_bytes=125952 peak_total_bytes=237556 maximum_standard_allocation_bytes=114688 active_tile_counts={1428, 1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:18:37.711803Z 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=1262 success=true 2026-09-06T11:18:42.516387Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=266344 estimated_exchange_cycles=79176 2026-09-06T11:18:42.604825Z 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 2026-09-06T11:18:42.607137Z 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=4895 success=true 2026-09-06T11:18:42.607170Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=26376 exchange_phases=4 2026-09-06T11:18:42.607426Z 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 2026-09-06T11:18:42.608183Z 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 2026-09-06T11:18:42.608497Z 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_k384_c48_r104_r106" target=ipu21 2026-09-06T11:18:42.670976Z 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/dbbc1b6f47491b27d15cce236bf67ff1bb5def91f36e88986784eabfdc06a6ee/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/dbbc1b6f47491b27d15cce236bf67ff1bb5def91f36e88986784eabfdc06a6ee/kernel.json 2026-09-06T11:18:42.671318Z 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_k384_c64_r104_r106" target=ipu21 2026-09-06T11:18:42.725375Z 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/f526d90239d55755f5eabb2a003c3ed312bc59537e8f1b24ae067b040c30d745/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/f526d90239d55755f5eabb2a003c3ed312bc59537e8f1b24ae067b040c30d745/kernel.json 2026-09-06T11:18:42.725722Z 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=117 success=true 2026-09-06T11:18:42.730186Z 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 2026-09-06T11:18:42.870646Z 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=140 success=true 2026-09-06T11:18:43.433494Z 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=16140 maximum_identical_destinations=1 2026-09-06T11:18:43.433545Z 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=2951 selected_horizon=2911 endpoint_lower_bound=2496 lower_bound_gap=415 selected_kind="critical-neighborhood" neighborhood_improvements=1 2026-09-06T11:18:43.444964Z 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=789 row_words=67 horizon=2911 2026-09-06T11:18:56.457900Z 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=5231 ordinary_horizon=45735 paired_horizon=38405 use_paired=true 2026-09-06T11:18:56.583850Z 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=7176 maximum_identical_destinations=1 2026-09-06T11:18:56.583903Z 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=38405 selected_horizon=38405 endpoint_lower_bound=23328 lower_bound_gap=15077 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:18:56.644751Z 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=727 row_words=349 horizon=38405 2026-09-06T11:19:03.832331Z 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 2026-09-06T11:19:03.832394Z 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 2026-09-06T11:19:03.895145Z 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 2026-09-06T11:19:03.921929Z 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=21051 success=true 2026-09-06T11:19:04.427756Z 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 2026-09-06T11:19:04.427818Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319176 profile_bytes=136 exchange_table_bytes=2400 host_code_bytes=2356 generated_code_bytes=1576 code_address=321536 standard_ranges=[(409600, 524288)] 2026-09-06T11:19:04.552710Z 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=124 success=true 2026-09-06T11:19:04.564042Z 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 2026-09-06T11:19:04.673717Z 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 2026-09-06T11:19:04.680887Z 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=16140 maximum_identical_destinations=1 2026-09-06T11:19:04.680912Z 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=2911 selected_horizon=2911 endpoint_lower_bound=2496 lower_bound_gap=415 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:19:04.691988Z 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=789 row_words=67 horizon=2911 2026-09-06T11:19:15.022989Z 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=5231 ordinary_horizon=45735 paired_horizon=38405 use_paired=true 2026-09-06T11:19:15.099704Z 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=7176 maximum_identical_destinations=1 2026-09-06T11:19:15.099749Z 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=38405 selected_horizon=38405 endpoint_lower_bound=23328 lower_bound_gap=15077 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:19:15.136940Z 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=727 row_words=349 horizon=38405 2026-09-06T11:19:22.759463Z 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 2026-09-06T11:19:22.759540Z 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 2026-09-06T11:19:22.822432Z 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 2026-09-06T11:19:22.849385Z 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=18296 success=true 2026-09-06T11:19:24.454189Z 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 2026-09-06T11:19:24.570912Z 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 2026-09-06T11:19:24.578761Z 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=16140 maximum_identical_destinations=1 2026-09-06T11:19:24.578788Z 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=2911 selected_horizon=2911 endpoint_lower_bound=2496 lower_bound_gap=415 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:19:24.589778Z 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=789 row_words=67 horizon=2911 2026-09-06T11:19:38.388604Z 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=5231 ordinary_horizon=40601 paired_horizon=29506 use_paired=true 2026-09-06T11:19:38.504353Z 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=7176 maximum_identical_destinations=1 2026-09-06T11:19:38.504396Z 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=29506 selected_horizon=29506 endpoint_lower_bound=23328 lower_bound_gap=6178 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:19:38.562895Z 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=727 row_words=349 horizon=29506 2026-09-06T11:19:44.700732Z 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 2026-09-06T11:19:44.700789Z 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 2026-09-06T11:19:44.740691Z 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 2026-09-06T11:19:44.769781Z 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=173256966144 baseline_cycles=61717 candidate_cycles=52818 row_bytes=2400 row_capacity=16320 accepted=true 2026-09-06T11:19:44.783731Z 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=21934 success=true 2026-09-06T11:19:44.952138Z 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 2026-09-06T11:19:44.965810Z 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=13 success=true 2026-09-06T11:19:45.076002Z 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=103 success=true 2026-09-06T11:19:45.150322Z INFO writing application package tiles=1472 2026-09-06T11:19:45.270184Z INFO application package written 2026-09-06T11:19:45.273086Z INFO reading application package 2026-09-06T11:19:45.383671Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:19:45.386889Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:19:45.386959Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:19:45.386966Z INFO initializing IPU device 2026-09-06T11:19:45.394773Z INFO IPU device initialized 2026-09-06T11:19:45.394784Z INFO replaying device configuration register_writes=2003 2026-09-06T11:19:45.403541Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:19:45.700233Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:19:48.056445Z INFO starting host exchange session startup_mark=23 2026-09-06T11:19:48.056538Z INFO attaching host exchange pages pages=2 2026-09-06T11:19:48.057239Z INFO host exchange session started 2026-09-06T11:19:48.057329Z INFO invoking streaming host exchange call call="initialize" command=0 phases=24 2026-09-06T11:19:48.066062Z INFO invoking streaming host exchange call call="run" command=0 phases=47 profile=artifacts/layout-sweep/up-01/execution.ipuprofile tiles=1472 samples=25816 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=114370560 weightBytes=19869696 cycles=215484 minimumTileCycles=192846 deviceMicroseconds=143.656 effectiveGemmTflops=100.644 peakTflops=282.624 efficiencyPercent=35.61 maximumAbsoluteError=0.009766 package=artifacts/layout-sweep/up-01/model.ipuexe tiles=1472 hardwareTest=PASS