2026-09-06T11:19:49.051067Z  INFO applied GEMM plan constraint source_operation=0 matching_plans=1 constraint=GemmPlanConstraint { source_operation: 0, orientation: Normal, row_partitions: 4, column_partitions: 90, 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 }
2026-09-06T11:19:49.491480Z  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 }
2026-09-06T11:19:50.037671Z  INFO retained operator-plan finalist finalist=0 values=7 operations=4 estimated_cycles=310377 estimated_exchange_cycles=123960 exchange_row_bytes=27236 peak_standard_bytes=141924 peak_interleaved_bytes=88064 peak_total_bytes=229988 maximum_standard_allocation_bytes=114688 active_tile_counts={1440} padding_group_counts={} conversion_sources=[Some(OperationId(2))]
2026-09-06T11:19:50.042958Z  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=1346 success=true
2026-09-06T11:19:53.286602Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}:ipu_codegen.package.phase{phase="select_finalist"}: selected analytical operator plan estimated_cycles=247983 estimated_exchange_cycles=70152
2026-09-06T11:19:53.373960Z  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
2026-09-06T11:19:53.376158Z  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=3333 success=true
2026-09-06T11:19:53.376192Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: lowered graph for package construction logical_shards=25956 exchange_phases=4
2026-09-06T11:19:53.376435Z  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
2026-09-06T11:19:53.377195Z  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
2026-09-06T11:19:53.377483Z  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=0 success=true
2026-09-06T11:19:53.380017Z  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
2026-09-06T11:19:53.517676Z  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=137 success=true
2026-09-06T11:19:54.066905Z  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
2026-09-06T11:19:54.066970Z  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=4174 selected_horizon=4168 endpoint_lower_bound=3408 lower_bound_gap=760 selected_kind="critical-neighborhood" neighborhood_improvements=1
2026-09-06T11:19:54.081911Z  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=1059 row_words=72 horizon=4168
2026-09-06T11:20:01.960139Z  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=3876 ordinary_horizon=75992 paired_horizon=65297 use_paired=true
2026-09-06T11:20:02.053089Z  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
2026-09-06T11:20:02.053130Z  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=65297 selected_horizon=65297 endpoint_lower_bound=23544 lower_bound_gap=41753 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T11:20:02.097784Z  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=1386 row_words=274 horizon=65297
2026-09-06T11:20:09.265478Z  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
2026-09-06T11:20:09.265542Z  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
2026-09-06T11:20:09.327973Z  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
2026-09-06T11:20:09.353712Z  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=15835 success=true
2026-09-06T11:20:09.970012Z  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
2026-09-06T11:20:09.970109Z  INFO ipu_codegen.package.build{tile_count=1472 operations=3}: allocated package support memory linked_end=317272 profile_bytes=132 exchange_table_bytes=2320 host_code_bytes=2412 generated_code_bytes=1512 code_address=319688 standard_ranges=[(409600, 524288)]
2026-09-06T11:20:10.096174Z  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=126 success=true
2026-09-06T11:20:10.105241Z  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
2026-09-06T11:20:10.207929Z  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
2026-09-06T11:20:10.215014Z  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
2026-09-06T11:20:10.215053Z  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=4168 selected_horizon=4168 endpoint_lower_bound=3408 lower_bound_gap=760 selected_kind="reused" neighborhood_improvements=0
2026-09-06T11:20:10.225294Z  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=1059 row_words=72 horizon=4168
2026-09-06T11:20:18.706959Z  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=3876 ordinary_horizon=75992 paired_horizon=65297 use_paired=true
2026-09-06T11:20:18.800886Z  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
2026-09-06T11:20:18.800925Z  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=65297 selected_horizon=65297 endpoint_lower_bound=23544 lower_bound_gap=41753 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T11:20:18.845761Z  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=1386 row_words=274 horizon=65297
2026-09-06T11:20:24.774679Z  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
2026-09-06T11:20:24.774744Z  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
2026-09-06T11:20:24.812267Z  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
2026-09-06T11:20:24.832012Z  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=14735 success=true
2026-09-06T11:20:26.154369Z  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
2026-09-06T11:20:26.261367Z  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
2026-09-06T11:20:26.268828Z  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
2026-09-06T11:20:26.268855Z  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=4168 selected_horizon=4168 endpoint_lower_bound=3408 lower_bound_gap=760 selected_kind="reused" neighborhood_improvements=0
2026-09-06T11:20:26.279585Z  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=1059 row_words=72 horizon=4168
2026-09-06T11:20:34.867725Z  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=3876 ordinary_horizon=50277 paired_horizon=40741 use_paired=true
2026-09-06T11:20:34.955048Z  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
2026-09-06T11:20:34.955094Z  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=40741 selected_horizon=40741 endpoint_lower_bound=23544 lower_bound_gap=17197 selected_kind="full-duplex" neighborhood_improvements=0
2026-09-06T11:20:35.000580Z  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=1386 row_words=276 horizon=40741
2026-09-06T11:20:42.729087Z  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
2026-09-06T11:20:42.729158Z  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
2026-09-06T11:20:42.793302Z  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
2026-09-06T11:20:42.832096Z  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=129706721280 score=115690733568 baseline_cycles=84500 candidate_cycles=59944 row_bytes=2272 row_capacity=16320 accepted=true
2026-09-06T11:20:42.849570Z  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=18017 success=true
2026-09-06T11:20:43.176045Z  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
2026-09-06T11:20:43.188780Z  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=12 success=true
2026-09-06T11:20:43.281314Z  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=86 success=true
2026-09-06T11:20:43.361914Z  INFO writing application package tiles=1472
2026-09-06T11:20:43.431867Z  INFO application package written
2026-09-06T11:20:43.434453Z  INFO reading application package
2026-09-06T11:20:43.544237Z  INFO application package read tiles=1472 compiler=0.1.0
2026-09-06T11:20:43.547139Z  INFO opening IPU device path="/dev/ipu0"
2026-09-06T11:20:43.547196Z  INFO IPU device attached path=/dev/ipu0
2026-09-06T11:20:43.547227Z  INFO initializing IPU device
2026-09-06T11:20:43.555198Z  INFO IPU device initialized
2026-09-06T11:20:43.555221Z  INFO replaying device configuration register_writes=2003
2026-09-06T11:20:43.564991Z  INFO loading application tile_count=1472 final_mark=23
2026-09-06T11:20:43.871325Z  INFO application loaded and initial synchronization completed tile_count=1472
2026-09-06T11:20:45.664814Z  INFO starting host exchange session startup_mark=23
2026-09-06T11:20:45.664891Z  INFO attaching host exchange pages pages=2
2026-09-06T11:20:45.665614Z  INFO host exchange session started
2026-09-06T11:20:45.665699Z  INFO invoking streaming host exchange call call="initialize" command=0 phases=14
2026-09-06T11:20:45.672880Z  INFO invoking streaming host exchange call call="run" command=0 phases=59
profile=artifacts/layout-sweep/up-02/execution.ipuprofile tiles=1472 samples=25676
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=151372800 weightBytes=19869696 cycles=201594 minimumTileCycles=191592 deviceMicroseconds=134.396 effectiveGemmTflops=107.579 peakTflops=282.624 efficiencyPercent=38.06 maximumAbsoluteError=0.011719
package=artifacts/layout-sweep/up-02/model.ipuexe tiles=1472 hardwareTest=PASS