2026-09-06T11:40:21.983988Z 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: 4, result_column_partitions: 1, output_column_block: 48, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:40:22.415879Z INFO applied GEMM plan constraint source_operation=2 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 2, orientation: Normal, row_partitions: 5, column_partitions: 36, inner_partitions: 8, result_row_partitions: 8, result_column_partitions: 1, output_column_block: 32, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:40:22.839575Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=342579 estimated_exchange_cycles=155320 exchange_row_bytes=34292 peak_standard_bytes=198132 peak_interleaved_bytes=95232 peak_total_bytes=267764 maximum_standard_allocation_bytes=163840 active_tile_counts={1440, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:40:22.844120Z 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=1198 success=true 2026-09-06T11:40:25.469074Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=247005 estimated_exchange_cycles=75880 2026-09-06T11:40:25.530118Z 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=144 2026-09-06T11:40:25.531350Z 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=2687 success=true 2026-09-06T11:40:25.531365Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=26408 exchange_phases=4 2026-09-06T11:40:25.531499Z 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 2026-09-06T11:40:25.531897Z 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 2026-09-06T11:40:25.532067Z 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_k544_c32_r146_r146" target=ipu21 2026-09-06T11:40:25.584989Z 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/1d1300be36354c3ae0fa7eac27cbb73411646b2e10e8ae7c0c02f15d990898c9/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/1d1300be36354c3ae0fa7eac27cbb73411646b2e10e8ae7c0c02f15d990898c9/kernel.json 2026-09-06T11:40:25.585366Z 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=53 success=true 2026-09-06T11:40:25.589312Z 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 2026-09-06T11:40:25.696497Z 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=107 success=true 2026-09-06T11:40:26.092255Z 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=15064 maximum_identical_destinations=1 2026-09-06T11:40:26.092288Z 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=4180 selected_horizon=3779 endpoint_lower_bound=3408 lower_bound_gap=371 selected_kind="matching-waves" neighborhood_improvements=0 2026-09-06T11:40:26.102769Z 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=690 row_words=75 horizon=3779 2026-09-06T11:40:43.303326Z 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=4330 ordinary_horizon=64791 paired_horizon=45532 use_paired=true 2026-09-06T11:40:43.436301Z 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=6244 maximum_identical_destinations=1 2026-09-06T11:40:43.436323Z 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=45532 selected_horizon=45532 endpoint_lower_bound=36312 lower_bound_gap=9220 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:40:43.500009Z 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=747 row_words=362 horizon=45532 2026-09-06T11:40:44.149037Z 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=22464 maximum_identical_destinations=1 2026-09-06T11:40:44.149061Z 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=3401 selected_horizon=2729 endpoint_lower_bound=2240 lower_bound_gap=489 selected_kind="critical-neighborhood" neighborhood_improvements=1 2026-09-06T11:40:44.161080Z 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=1149 row_words=100 horizon=2729 2026-09-06T11:40:44.165759Z 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=18469 success=true 2026-09-06T11:40:44.531661Z 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=4 success=true 2026-09-06T11:40:44.531687Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=318224 profile_bytes=196 exchange_table_bytes=2012 host_code_bytes=2516 generated_code_bytes=1876 code_address=320744 standard_ranges=[(409600, 524288)] 2026-09-06T11:40:44.632962Z 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=101 success=true 2026-09-06T11:40:44.640398Z 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 2026-09-06T11:40:44.729392Z 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 2026-09-06T11:40:44.735359Z 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=15064 maximum_identical_destinations=1 2026-09-06T11:40:44.735371Z 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=3779 selected_horizon=3779 endpoint_lower_bound=3408 lower_bound_gap=371 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:40:44.745259Z 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=690 row_words=75 horizon=3779 2026-09-06T11:41:03.310645Z 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=4330 ordinary_horizon=64791 paired_horizon=45532 use_paired=true 2026-09-06T11:41:03.443570Z 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=6244 maximum_identical_destinations=1 2026-09-06T11:41:03.443602Z 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=45532 selected_horizon=45532 endpoint_lower_bound=36312 lower_bound_gap=9220 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:41:03.506554Z 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=747 row_words=362 horizon=45532 2026-09-06T11:41:03.690075Z 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=3 2026-09-06T11:41:03.699925Z 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=22464 maximum_identical_destinations=1 2026-09-06T11:41:03.699942Z 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=2729 selected_horizon=2729 endpoint_lower_bound=2240 lower_bound_gap=489 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:41:03.720260Z 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=1149 row_words=100 horizon=2729 2026-09-06T11:41:03.726450Z 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=19093 success=true 2026-09-06T11:41:05.091491Z 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 2026-09-06T11:41:05.186794Z 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 2026-09-06T11:41:05.193288Z 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=15064 maximum_identical_destinations=1 2026-09-06T11:41:05.193301Z 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=3779 selected_horizon=3779 endpoint_lower_bound=3408 lower_bound_gap=371 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:41:05.203741Z 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=690 row_words=75 horizon=3779 2026-09-06T11:41:23.887352Z 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=4330 ordinary_horizon=56467 paired_horizon=42430 use_paired=true 2026-09-06T11:41:24.019667Z 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=6244 maximum_identical_destinations=1 2026-09-06T11:41:24.019693Z 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=42430 selected_horizon=42430 endpoint_lower_bound=36312 lower_bound_gap=6118 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:41:24.087924Z 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=978 row_words=363 horizon=42430 2026-09-06T11:41:24.270380Z 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=3 2026-09-06T11:41:24.280062Z 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=22464 maximum_identical_destinations=1 2026-09-06T11:41:24.280075Z 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=2729 selected_horizon=2729 endpoint_lower_bound=2240 lower_bound_gap=489 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:41:24.300898Z 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=1149 row_words=100 horizon=2729 2026-09-06T11:41:24.314851Z 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=563315081216 score=166408224768 baseline_cycles=59088 candidate_cycles=55986 row_bytes=2004 row_capacity=16320 accepted=true 2026-09-06T11:41:24.324853Z 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=20598 success=true 2026-09-06T11:41:24.636408Z 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=9 success=true 2026-09-06T11:41:24.647087Z 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=10 success=true 2026-09-06T11:41:24.745821Z 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=94 success=true 2026-09-06T11:41:24.811774Z INFO writing application package tiles=1472 2026-09-06T11:41:24.871622Z INFO application package written 2026-09-06T11:41:24.873302Z INFO reading application package 2026-09-06T11:41:24.975250Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:41:24.977564Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:41:24.977579Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:41:24.977584Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:41:24.977617Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:41:24.977620Z INFO initializing IPU device 2026-09-06T11:41:24.985532Z INFO IPU device initialized 2026-09-06T11:41:24.985536Z INFO replaying device configuration register_writes=2003 2026-09-06T11:41:24.994394Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:41:25.368757Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:41:27.933891Z INFO starting host exchange session startup_mark=23 2026-09-06T11:41:27.933920Z INFO attaching host exchange pages pages=2 2026-09-06T11:41:27.934389Z INFO host exchange session started 2026-09-06T11:41:27.934432Z INFO invoking streaming host exchange call call="initialize" command=0 phases=18 2026-09-06T11:41:27.940084Z INFO invoking streaming host exchange call call="run" command=0 phases=59 profile=artifacts/layout-sweep/down-04/execution.ipuprofile tiles=1472 samples=29205 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=19943424 cycles=203400 minimumTileCycles=194160 deviceMicroseconds=135.600 effectiveGemmTflops=106.623 peakTflops=282.624 efficiencyPercent=37.73 maximumAbsoluteError=0.011719 package=artifacts/layout-sweep/down-04/model.ipuexe tiles=1472 hardwareTest=PASS