1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
|
# Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of NVIDIA CORPORATION nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
import NvRules
from RequestedMetrics import MetricRequest, RequestedMetricsParser
requested_metrics = [
MetricRequest("smsp__maximum_warps_avg_per_active_cycle", "theoretical_warps"),
MetricRequest("smsp__warps_active.avg.peak_sustained", "max_warps"),
MetricRequest("launch__occupancy_limit_blocks"),
MetricRequest("launch__occupancy_limit_registers"),
MetricRequest("launch__occupancy_limit_shared_mem"),
MetricRequest("launch__occupancy_limit_warps"),
]
def get_identifier():
return "TheoreticalOccupancy"
def get_name():
return "Theoretical Occupancy"
def get_description():
return "Analysis of Theoretical Occupancy and its Limiters"
def get_section_identifier():
return "Occupancy"
def get_parent_rules_identifiers():
return ["IssueSlotUtilization"]
def get_estimated_speedup(parent_weights, metrics):
theoretical_warps = metrics["theoretical_warps"].value()
max_warps = metrics["max_warps"].value()
improvement_local = 1 - theoretical_warps / max_warps
parent_speedup_name = "issue_slot_util_speedup_normalized"
if parent_speedup_name in parent_weights:
speedup_type = NvRules.IFrontend.SpeedupType_GLOBAL
improvement_global = min(parent_weights[parent_speedup_name], improvement_local)
improvement_percent = improvement_global * 100
else:
speedup_type = NvRules.IFrontend.SpeedupType_LOCAL
improvement_percent = improvement_local * 100
return speedup_type, improvement_percent
def apply(handle):
ctx = NvRules.get_context(handle)
action = ctx.range_by_idx(0).action_by_idx(0)
fe = ctx.frontend()
metrics = RequestedMetricsParser(handle, action).parse(requested_metrics)
parent_weights = fe.receive_dict_from_parent("IssueSlotUtilization")
theoretical_warps = metrics["theoretical_warps"].value()
max_warps = metrics["max_warps"].value()
theoretical_warps_pct_of_peak = (theoretical_warps / max_warps) * 100
low_theoretical_threshold = 80
if theoretical_warps_pct_of_peak < low_theoretical_threshold:
message = "The {:.2f} theoretical warps per scheduler this kernel can issue according to its occupancy are below the hardware maximum of {}.".format(
theoretical_warps, int(max_warps)
)
limit_types = {
"blocks": "the number of blocks that can fit on the SM",
"registers": "the number of required registers",
"shared_mem": "the required amount of shared memory",
"warps": "the number of warps within each block",
}
limiters = []
for limiter in limit_types:
limit_value = metrics[f"launch__occupancy_limit_{limiter}"].value()
limit_msg = limit_types[limiter]
limiters.append((limiter, limit_value, limit_msg))
sorted_limiters = sorted(limiters, key=lambda limit: limit[1])
last_limiter = -1
for limiter in sorted_limiters:
value = limiter[1]
if last_limiter == -1 or value == last_limiter:
message += " This kernel's theoretical occupancy ({:.1f}%) is limited by {}.".format(
theoretical_warps_pct_of_peak, limiter[2]
)
last_limiter = value
msg_id = fe.message(NvRules.IFrontend.MsgType_MSG_OPTIMIZATION, message)
speedup_type, speedup_value = get_estimated_speedup(parent_weights, metrics)
fe.speedup(msg_id, speedup_type, speedup_value)
fe.focus_metric(
msg_id,
metrics["theoretical_warps"].name(),
theoretical_warps,
NvRules.IFrontend.Severity_SEVERITY_HIGH,
"Increase the theoretical number of warps per schedule that can be issued",
)
|
