2026-09-06T11:42:59.380079Z 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 } 2026-09-06T11:42:59.812028Z INFO applied GEMM plan constraint source_operation=2 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 2, orientation: Normal, row_partitions: 8, column_partitions: 36, inner_partitions: 5, result_row_partitions: 5, result_column_partitions: 1, output_column_block: 32, weight_memory_class: Ipu21Interleaved, reduction_staging: Complete, local_weight_staging: Direct } 2026-09-06T11:43:00.095130Z INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=361147 estimated_exchange_cycles=165720 exchange_row_bytes=36632 peak_standard_bytes=200472 peak_interleaved_bytes=115712 peak_total_bytes=311064 maximum_standard_allocation_bytes=163840 active_tile_counts={1440, 1472} padding_group_counts={} conversion_sources=[Some(OperationId(2))] 2026-09-06T11:43:00.098796Z 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=1007 success=true 2026-09-06T11:43:05.947855Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=257084 estimated_exchange_cycles=79112 2026-09-06T11:43:06.079240Z 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=252 2026-09-06T11:43:06.082194Z 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=5983 success=true 2026-09-06T11:43:06.082244Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=26300 exchange_phases=4 2026-09-06T11:43:06.082502Z 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 2026-09-06T11:43:06.084104Z 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=1 success=true 2026-09-06T11:43:06.084510Z 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_k864_c32_r90_r92" target=ipu21 2026-09-06T11:43:06.147968Z 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/5ed027ebc3d6b6ffbbcc23d93ba9903862e990eff5728a9bd8ab8f31d0998ab4/kernel.o metadata=/srv/home/.cache/ipu-stack/kernels/5ed027ebc3d6b6ffbbcc23d93ba9903862e990eff5728a9bd8ab8f31d0998ab4/kernel.json 2026-09-06T11:43:06.148275Z 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=64 success=true 2026-09-06T11:43:06.152280Z 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 2026-09-06T11:43:06.286674Z 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=134 success=true 2026-09-06T11:43:06.625046Z 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 2026-09-06T11:43:06.625075Z 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 2026-09-06T11:43:06.635923Z 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 2026-09-06T11:43:20.198856Z 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=3473 ordinary_horizon=58812 paired_horizon=51572 use_paired=true 2026-09-06T11:43:20.268785Z 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 2026-09-06T11:43:20.268804Z 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=51572 selected_horizon=51572 endpoint_lower_bound=43632 lower_bound_gap=7940 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:43:20.303897Z 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=1259 row_words=399 horizon=51572 2026-09-06T11:43:20.862355Z 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=12960 maximum_identical_destinations=1 2026-09-06T11:43:20.862388Z 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=1771 selected_horizon=1732 endpoint_lower_bound=1408 lower_bound_gap=324 selected_kind="critical-neighborhood" neighborhood_improvements=1 2026-09-06T11:43:20.873771Z 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=507 row_words=62 horizon=1732 2026-09-06T11:43:20.878164Z 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=14591 success=true 2026-09-06T11:43:21.491864Z 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=9 success=true 2026-09-06T11:43:21.491898Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=319152 profile_bytes=276 exchange_table_bytes=1944 host_code_bytes=2836 generated_code_bytes=2356 code_address=321992 standard_ranges=[(409600, 524288)] 2026-09-06T11:43:21.591613Z 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=99 success=true 2026-09-06T11:43:21.599098Z 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 2026-09-06T11:43:21.690908Z 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 2026-09-06T11:43:21.697162Z 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 2026-09-06T11:43:21.697175Z 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 2026-09-06T11:43:21.707822Z 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 2026-09-06T11:43:37.635433Z 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=3473 ordinary_horizon=58812 paired_horizon=51572 use_paired=true 2026-09-06T11:43:37.750372Z 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 2026-09-06T11:43:37.750398Z 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=51572 selected_horizon=51572 endpoint_lower_bound=43632 lower_bound_gap=7940 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:43:37.807903Z 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=1259 row_words=399 horizon=51572 2026-09-06T11:43:37.914731Z 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 2026-09-06T11:43:37.920323Z 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=12960 maximum_identical_destinations=1 2026-09-06T11:43:37.920338Z 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=1732 selected_horizon=1732 endpoint_lower_bound=1408 lower_bound_gap=324 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:43:37.931762Z 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=507 row_words=62 horizon=1732 2026-09-06T11:43:37.935336Z 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=16343 success=true 2026-09-06T11:43:39.301869Z 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 2026-09-06T11:43:39.393791Z 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 2026-09-06T11:43:39.400098Z 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 2026-09-06T11:43:39.400112Z 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 2026-09-06T11:43:39.410549Z 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 2026-09-06T11:43:55.046005Z 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=3473 ordinary_horizon=58812 paired_horizon=50515 use_paired=true 2026-09-06T11:43:55.160396Z 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 2026-09-06T11:43:55.160424Z 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=50515 selected_horizon=50515 endpoint_lower_bound=43632 lower_bound_gap=6883 selected_kind="full-duplex" neighborhood_improvements=0 2026-09-06T11:43:55.216573Z 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=1259 row_words=401 horizon=50515 2026-09-06T11:43:55.323959Z 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 2026-09-06T11:43:55.329634Z 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=12960 maximum_identical_destinations=1 2026-09-06T11:43:55.329648Z 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=1732 selected_horizon=1732 endpoint_lower_bound=1408 lower_bound_gap=324 selected_kind="reused" neighborhood_improvements=0 2026-09-06T11:43:55.340993Z 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=507 row_words=62 horizon=1732 2026-09-06T11:43:55.351099Z INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="optimize_exchange_placement"}: evaluated exchange placement finalist offset=8192 baseline_score=890600030208 score=340223459328 baseline_cycles=64131 candidate_cycles=63074 row_bytes=1944 row_capacity=16320 accepted=true 2026-09-06T11:43:55.361474Z 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=17426 success=true 2026-09-06T11:43:55.676553Z 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=16 success=true 2026-09-06T11:43:55.688004Z 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 2026-09-06T11:43:55.787962Z 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=95 success=true 2026-09-06T11:43:55.838278Z INFO writing application package tiles=1472 2026-09-06T11:43:55.888946Z INFO application package written 2026-09-06T11:43:55.890899Z INFO reading application package 2026-09-06T11:43:55.995061Z INFO application package read tiles=1472 compiler=0.1.0 2026-09-06T11:43:55.997282Z INFO waiting for hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:43:55.997301Z INFO acquired hardware ownership path=/srv/home/gc-sdk/ipu-stack/artifacts/layout-sweep/device.lock 2026-09-06T11:43:55.997306Z INFO opening IPU device path="/dev/ipu0" 2026-09-06T11:43:55.997334Z INFO IPU device attached path=/dev/ipu0 2026-09-06T11:43:55.997337Z INFO initializing IPU device 2026-09-06T11:43:56.005608Z INFO IPU device initialized 2026-09-06T11:43:56.005659Z INFO replaying device configuration register_writes=2003 2026-09-06T11:43:56.017986Z INFO loading application tile_count=1472 final_mark=23 2026-09-06T11:43:56.365019Z INFO application loaded and initial synchronization completed tile_count=1472 2026-09-06T11:43:59.078307Z INFO starting host exchange session startup_mark=23 2026-09-06T11:43:59.078338Z INFO attaching host exchange pages pages=2 2026-09-06T11:43:59.078986Z INFO host exchange session started 2026-09-06T11:43:59.079026Z INFO invoking streaming host exchange call call="initialize" command=0 phases=28 2026-09-06T11:43:59.086996Z INFO invoking streaming host exchange call call="run" command=0 phases=59 profile=artifacts/layout-sweep/down-14/execution.ipuprofile tiles=1472 samples=29298 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=19869696 cycles=219228 minimumTileCycles=210918 deviceMicroseconds=146.152 effectiveGemmTflops=98.925 peakTflops=282.624 efficiencyPercent=35.00 maximumAbsoluteError=0.011719 package=artifacts/layout-sweep/down-14/model.ipuexe tiles=1472 hardwareTest=PASS