2026-09-06T11:37:19.684972Z INFO applied GEMM plan constraint source_operation=0 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 1, column_partitions: 135, inner_partitions: 9, result_row_partitions: 9, result_column_partitions: 1, output_column_block: 32, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:37:20.114012Z 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 } 2026-09-06T11:37:20.655377Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=338447 estimated_exchange_cycles=124000 exchange_row_bytes=27376 peak_standard_bytes=223984 peak_interleaved_bytes=101376 peak_total_bytes=325360 maximum_standard_allocation_bytes=196608 active_tile_counts={1215, 1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:37:20.664900Z 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=1271 success=true 2026-09-06T11:37:22.775735Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=278309 estimated_exchange_cycles=69520 2026-09-06T11:37:22.859327Z 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 2026-09-06T11:37:22.860955Z 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=2195 success=true 2026-09-06T11:37:22.860972Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=23031 exchange_phases=3 2026-09-06T11:37:22.861166Z 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 2026-09-06T11:37:22.861662Z 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 2026-09-06T11:37:22.861809Z 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_k128_c16_r730_r730" target=ipu21 2026-09-06T11:37:22.913131Z 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/fe7996963e7c294fe6685eb8e54174b4b20f94b9730008104b2ed5f727d3598a/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/fe7996963e7c294fe6685eb8e54174b4b20f94b9730008104b2ed5f727d3598a/kernel.json 2026-09-06T11:37:22.913497Z 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_k128_c32_r730_r730" target=ipu21 2026-09-06T11:37:22.957826Z 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/904de571caa8f8fdb52767f616a47faa5c3b9c4c8930cf2ab3c89614990429f5/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/904de571caa8f8fdb52767f616a47faa5c3b9c4c8930cf2ab3c89614990429f5/kernel.json 2026-09-06T11:37:22.958375Z 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 2026-09-06T11:37:22.960869Z 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 2026-09-06T11:37:23.071476Z 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=110 success=true 2026-09-06T11:37:23.599521Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: prepared large physical exchange phase phase=0 transfers=19368 maximum_identical_destinations=1 2026-09-06T11:37:23.599545Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: optimized physical exchange schedule phase=0 initial_horizon=13722 selected_horizon=10712 endpoint_lower_bound=10496 lower_bound_gap=216 selected_kind="matching-waves" neighborhood_improvements=0 2026-09-06T11:37:23.618368Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges_provisional"}: finished large physical exchange phase phase=0 tile=767 row_words=100 horizon=10712 2026-09-06T11:37:25.209760Z 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=1 candidates=3048 ordinary_horizon=37408 paired_horizon=28389 use_paired=true 2026-09-06T11:37:25.260070Z 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=3948 maximum_identical_destinations=1 2026-09-06T11:37:25.260090Z 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=28389 selected_horizon=28389 endpoint_lower_bound=25920 lower_bound_gap=2469 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:37:25.296253Z 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=248 row_words=204 horizon=28389 2026-09-06T11:37:31.391285Z 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=63720 maximum_identical_destinations=1 2026-09-06T11:37:31.391332Z 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=7991 selected_horizon=7991 endpoint_lower_bound=4704 lower_bound_gap=3287 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:37:31.453480Z 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=1058 row_words=294 horizon=7991 2026-09-06T11:37:31.475573Z 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=8404 success=true 2026-09-06T11:37:32.273751Z 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 2026-09-06T11:37:32.273789Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=317272 profile_bytes=128 exchange_table_bytes=2080 host_code_bytes=3812 generated_code_bytes=1784 code_address=321088 standard_ranges=[(409600, 524288)] 2026-09-06T11:37:32.368351Z 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=94 success=true 2026-09-06T11:37:32.516978Z 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 2026-09-06T11:37:32.524679Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: prepared large physical exchange phase phase=0 transfers=19368 maximum_identical_destinations=1 2026-09-06T11:37:32.524693Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: optimized physical exchange schedule phase=0 initial_horizon=10712 selected_horizon=10712 endpoint_lower_bound=10496 lower_bound_gap=216 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:37:32.543032Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="lower_exchanges"}: finished large physical exchange phase phase=0 tile=767 row_words=100 horizon=10712 2026-09-06T11:37:33.065989Z 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 2026-09-06T11:37:33.082691Z 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=3948 maximum_identical_destinations=1 2026-09-06T11:37:33.082707Z 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=28389 selected_horizon=28389 endpoint_lower_bound=25920 lower_bound_gap=2469 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:37:33.108484Z 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=248 row_words=204 horizon=28389 2026-09-06T11:37:39.856092Z 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=63720 maximum_identical_destinations=1 2026-09-06T11:37:39.856153Z 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=7991 selected_horizon=7991 endpoint_lower_bound=4704 lower_bound_gap=3287 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:37:39.936260Z 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=1058 row_words=294 horizon=7991 2026-09-06T11:37:39.962085Z 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=7593 success=true 2026-09-06T11:37:41.306712Z 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 2026-09-06T11:37:41.315059Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: prepared large physical exchange phase phase=0 transfers=19368 maximum_identical_destinations=1 2026-09-06T11:37:41.315076Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: optimized physical exchange schedule phase=0 initial_horizon=10712 selected_horizon=10712 endpoint_lower_bound=10496 lower_bound_gap=216 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:37:41.333865Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: finished large physical exchange phase phase=0 tile=767 row_words=100 horizon=10712 2026-09-06T11:37:43.351806Z 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=1 candidates=3048 ordinary_horizon=37408 paired_horizon=28389 use_paired=true 2026-09-06T11:37:43.403258Z 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=3948 maximum_identical_destinations=1 2026-09-06T11:37:43.403279Z 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=28389 selected_horizon=28389 endpoint_lower_bound=25920 lower_bound_gap=2469 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:37:43.430769Z 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=248 row_words=201 horizon=28389 2026-09-06T11:37:50.329905Z 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=63720 maximum_identical_destinations=1 2026-09-06T11:37:50.329947Z 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=7991 selected_horizon=7991 endpoint_lower_bound=4704 lower_bound_gap=3287 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:37:50.377450Z 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=1058 row_words=294 horizon=7991 2026-09-06T11:37:50.403695Z 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=121407897600 score=84360462336 baseline_cycles=47092 candidate_cycles=47092 row_bytes=2080 row_capacity=16320 accepted=false 2026-09-06T11:37:50.414836Z 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=10452 success=true 2026-09-06T11:37:50.671160Z 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 2026-09-06T11:37:50.682809Z 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 2026-09-06T11:37:50.768924Z 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=82 success=true 2026-09-06T11:37:50.867037Z INFO writing application package tiles=1472 2026-09-06T11:37:50.956673Z INFO application package written 2026-09-06T11:37:50.958390Z INFO reading application package 2026-09-06T11:37:51.056242Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:37:51.058805Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:37:51.058817Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:37:51.058826Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:37:51.058856Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:37:51.058858Z INFO initializing IPU device 2026-09-06T11:37:51.066688Z INFO IPU device initialized 2026-09-06T11:37:51.066716Z INFO replaying device configuration register_writes=2003 2026-09-06T11:37:51.077125Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:37:51.411955Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:37:54.772014Z INFO starting host exchange session startup_mark=23 2026-09-06T11:37:54.772045Z INFO attaching host exchange pages pages=2 2026-09-06T11:37:54.772775Z INFO host exchange session started 2026-09-06T11:37:54.772824Z INFO invoking streaming host exchange call call="initialize" command=0 phases=18 2026-09-06T11:37:54.780028Z INFO invoking streaming host exchange call call="run" command=0 phases=99 profile=artifacts/layout-sweep/up-10/execution.ipuprofile tiles=1472 samples=22845 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=227059200 weightBytes=19869696 cycles=211032 minimumTileCycles=193908 deviceMicroseconds=140.688 effectiveGemmTflops=102.767 peakTflops=282.624 efficiencyPercent=36.36 maximumAbsoluteError=0.011719 package=artifacts/layout-sweep/up-10/model.ipuexe tiles=1472 hardwareTest=PASS