2026-09-06T11:28:00.044745Z INFO applied GEMM plan constraint source_operation=0 matching_plans=2 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 2, column_partitions: 30, inner_partitions: 24, result_row_partitions: 24, result_column_partitions: 1, output_column_block: 144, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:28:00.426957Z 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 } 2026-09-06T11:28:00.966612Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=445256 estimated_exchange_cycles=173240 exchange_row_bytes=38324 peak_standard_bytes=153524 peak_interleaved_bytes=172544 peak_total_bytes=316852 maximum_standard_allocation_bytes=115200 active_tile_counts={1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:28:00.973823Z 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=1219 success=true 2026-09-06T11:28:03.645370Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=403867 estimated_exchange_cycles=131672 2026-09-06T11:28:03.734000Z 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 2026-09-06T11:28:03.735889Z 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=2761 success=true 2026-09-06T11:28:03.735904Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=25866 exchange_phases=4 2026-09-06T11:28:03.736100Z 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 2026-09-06T11:28:03.736652Z 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 2026-09-06T11:28:03.736800Z 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_k48_c128_r364_r366" target=ipu21 2026-09-06T11:28:03.787346Z 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/f9512ad903bdd5955126c605539891825b95bcc804327a0f3abbe6dcbf9fe653/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/f9512ad903bdd5955126c605539891825b95bcc804327a0f3abbe6dcbf9fe653/kernel.json 2026-09-06T11:28:03.787676Z 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_k48_c144_r364_r366" target=ipu21 2026-09-06T11:28:03.831927Z 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/f110eed7c6de2e96c9fef4161482fd5a46120bbacc874c06b4782c34624940fa/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/f110eed7c6de2e96c9fef4161482fd5a46120bbacc874c06b4782c34624940fa/kernel.json 2026-09-06T11:28:03.832443Z 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=95 success=true 2026-09-06T11:28:03.835175Z 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 2026-09-06T11:28:03.952082Z 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 2026-09-06T11:28:49.891970Z 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=303701 maximum_identical_destinations=1 2026-09-06T11:28:49.892014Z 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=49986 selected_horizon=49986 endpoint_lower_bound=27648 lower_bound_gap=22338 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:28:50.196816Z 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=688 row_words=1246 horizon=49986 2026-09-06T11:29:33.487520Z 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=6795 ordinary_horizon=47928 paired_horizon=49994 use_paired=false 2026-09-06T11:29:33.709700Z 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=14170 maximum_identical_destinations=1 2026-09-06T11:29:33.709724Z 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=47928 selected_horizon=47928 endpoint_lower_bound=33408 lower_bound_gap=14520 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:29:33.824410Z 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=875 row_words=582 horizon=47928 2026-09-06T11:29:38.975916Z 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 2026-09-06T11:29:38.975945Z 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 2026-09-06T11:29:39.016301Z 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 2026-09-06T11:29:39.030864Z 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=95078 success=true 2026-09-06T11:29:39.270026Z 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 2026-09-06T11:29:39.270047Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319128 profile_bytes=132 exchange_table_bytes=7876 host_code_bytes=1300 generated_code_bytes=1484 code_address=320432 standard_ranges=[(409600, 524288)] 2026-09-06T11:29:39.367311Z 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=97 success=true 2026-09-06T11:29:39.371279Z 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 2026-09-06T11:29:45.918078Z 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 2026-09-06T11:29:46.136498Z 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=303701 maximum_identical_destinations=1 2026-09-06T11:29:46.136534Z 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=49986 selected_horizon=49986 endpoint_lower_bound=27648 lower_bound_gap=22338 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:29:46.634862Z 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=688 row_words=1246 horizon=49986 2026-09-06T11:30:39.314100Z 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=6795 ordinary_horizon=47928 paired_horizon=49994 use_paired=false 2026-09-06T11:30:39.546615Z 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=14170 maximum_identical_destinations=1 2026-09-06T11:30:39.546642Z 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=47928 selected_horizon=47928 endpoint_lower_bound=33408 lower_bound_gap=14520 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:30:39.659155Z 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=875 row_words=582 horizon=47928 2026-09-06T11:30:46.242332Z 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 2026-09-06T11:30:46.242369Z 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 2026-09-06T11:30:46.309617Z 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 2026-09-06T11:30:46.331538Z 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=66964 success=true 2026-09-06T11:30:48.299669Z 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 2026-09-06T11:30:54.687623Z 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 2026-09-06T11:30:54.848297Z 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=303701 maximum_identical_destinations=1 2026-09-06T11:30:54.848333Z 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=49986 selected_horizon=49986 endpoint_lower_bound=27648 lower_bound_gap=22338 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:30:55.359786Z 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=688 row_words=1246 horizon=49986 2026-09-06T11:31:39.325062Z 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=6795 ordinary_horizon=39995 paired_horizon=43302 use_paired=false 2026-09-06T11:31:39.483564Z 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=14170 maximum_identical_destinations=1 2026-09-06T11:31:39.483586Z 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=39995 selected_horizon=39995 endpoint_lower_bound=33408 lower_bound_gap=6587 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:31:39.555963Z 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=875 row_words=592 horizon=39995 2026-09-06T11:31:44.011109Z 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 2026-09-06T11:31:44.011139Z 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 2026-09-06T11:31:44.059520Z 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 2026-09-06T11:31:44.111999Z 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=131134685184 score=121848496128 baseline_cycles=109575 candidate_cycles=101642 row_bytes=7868 row_capacity=16320 accepted=true 2026-09-06T11:31:44.155952Z 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=57824 success=true 2026-09-06T11:31:44.318418Z 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=8 success=true 2026-09-06T11:31:44.329913Z 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 2026-09-06T11:31:44.399997Z 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=65 success=true 2026-09-06T11:31:44.454165Z INFO writing application package tiles=1472 2026-09-06T11:31:44.520081Z INFO application package written 2026-09-06T11:31:44.520220Z INFO reading application package 2026-09-06T11:31:44.624149Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:31:44.627253Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:31:44.627265Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:31:44.627269Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:31:44.627293Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:31:44.627296Z INFO initializing IPU device 2026-09-06T11:31:44.635097Z INFO IPU device initialized 2026-09-06T11:31:44.635100Z INFO replaying device configuration register_writes=2003 2026-09-06T11:31:44.641335Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:31:44.890934Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:31:45.864825Z INFO starting host exchange session startup_mark=23 2026-09-06T11:31:45.864875Z INFO attaching host exchange pages pages=2 2026-09-06T11:31:45.865579Z INFO host exchange session started 2026-09-06T11:31:45.865636Z INFO invoking streaming host exchange call call="initialize" command=0 phases=14 2026-09-06T11:31:45.870831Z INFO invoking streaming host exchange call call="run" command=0 phases=25 profile=artifacts/layout-sweep/up-08/execution.ipuprofile tiles=1472 samples=25444 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=50457600 weightBytes=19869696 cycles=268488 minimumTileCycles=264000 deviceMicroseconds=178.992 effectiveGemmTflops=80.775 peakTflops=282.624 efficiencyPercent=28.58 maximumAbsoluteError=0.009766 package=artifacts/layout-sweep/up-08/model.ipuexe tiles=1472 hardwareTest=PASS