118 lines
2.2 KiB
YAML
118 lines
2.2 KiB
YAML
|
|
# muh schema for histogram
|
||
|
|
# Auto-extracted from cub/cub/device/dispatch/tuning/tuning_histogram.cuh
|
||
|
|
# Generated by muh/extract.py
|
||
|
|
|
||
|
|
algorithm: histogram
|
||
|
|
source: cub/cub/device/dispatch/tuning/tuning_histogram.cuh
|
||
|
|
parameters:
|
||
|
|
threads_per_block:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 32
|
||
|
|
- 1024
|
||
|
|
step: 32
|
||
|
|
pixels_per_thread:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
vec_size:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 8
|
||
|
|
step: 1
|
||
|
|
load_algorithm:
|
||
|
|
type: enum
|
||
|
|
values:
|
||
|
|
- BLOCK_LOAD_DIRECT
|
||
|
|
- BLOCK_LOAD_VECTORIZE
|
||
|
|
- BLOCK_LOAD_TRANSPOSE
|
||
|
|
- BLOCK_LOAD_WARP_TRANSPOSE
|
||
|
|
- BLOCK_LOAD_WARP_TRANSPOSE_TIMESLICED
|
||
|
|
- BLOCK_LOAD_STRIPED
|
||
|
|
load_modifier:
|
||
|
|
type: enum
|
||
|
|
values:
|
||
|
|
- LOAD_DEFAULT
|
||
|
|
- LOAD_CA
|
||
|
|
- LOAD_CG
|
||
|
|
- LOAD_CS
|
||
|
|
- LOAD_CV
|
||
|
|
- LOAD_LDG
|
||
|
|
init_kernel_pdl_trigger_max_bins:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
sample_size:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
counter_size:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
sample_size_bytes:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
num_channels:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
num_active_channels:
|
||
|
|
type: int
|
||
|
|
range:
|
||
|
|
- 1
|
||
|
|
- 1024
|
||
|
|
step: 1
|
||
|
|
note: unknown_range
|
||
|
|
reference_tunings:
|
||
|
|
sm90:
|
||
|
|
-
|
||
|
|
threads: 768
|
||
|
|
items: 12
|
||
|
|
load_modifier: LOAD_LDG
|
||
|
|
load_algorithm: BLOCK_LOAD_DIRECT
|
||
|
|
-
|
||
|
|
threads: 960
|
||
|
|
items: 10
|
||
|
|
load_modifier: LOAD_DEFAULT
|
||
|
|
load_algorithm: BLOCK_LOAD_DIRECT
|
||
|
|
sm100:
|
||
|
|
-
|
||
|
|
items: 12
|
||
|
|
threads: 928
|
||
|
|
load_modifier: LOAD_CA
|
||
|
|
load_algorithm: BLOCK_LOAD_DIRECT
|
||
|
|
vec_size: 1
|
||
|
|
-
|
||
|
|
items: 12
|
||
|
|
threads: 448
|
||
|
|
load_modifier: LOAD_LDG
|
||
|
|
load_algorithm: BLOCK_LOAD_DIRECT
|
||
|
|
vec_size: 1
|
||
|
|
init_kernel_pdl_trigger_max_bins: 2048
|
||
|
|
bi_v100:
|
||
|
|
status: pending_benchmark
|
||
|
|
note: Run muh benchmark on Iluvatar BI-V100 to fill these values
|
||
|
|
threads_per_block: TBD
|
||
|
|
items_per_thread: TBD
|