Skip to content

Commit c00832c

Browse files
Sean1783Sean Archer
andauthored
Add GPU operator MIG support with NVIDIA license notice (#310)
Co-authored-by: Sean Archer <[email protected]>
1 parent bc62d4f commit c00832c

39 files changed

+856
-1
lines changed

NOTICE

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1,3 @@
1-
Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
1+
Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
2+
3+
This software contains source code provided by NVIDIA Corporation.

helm_chart/HyperPodHelmChart/Chart.yaml

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -88,3 +88,7 @@ dependencies:
8888
version: "0.1.0"
8989
repository: "file://charts/hyperpod-patching"
9090
condition: hyperpod-patching.enabled
91+
- name: gpu-operator
92+
version: "0.1.0"
93+
repository: "file://charts/gpu-operator"
94+
condition: gpu-operator.enabled
Lines changed: 30 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,30 @@
1+
apiVersion: v2
2+
name: gpu-operator
3+
description: A Helm chart for Kubernetes
4+
5+
# A chart can be either an 'application' or a 'library' chart.
6+
#
7+
# Application charts are a collection of templates that can be packaged into versioned archives
8+
# to be deployed.
9+
#
10+
# Library charts provide useful utilities or functions for the chart developer. They're included as
11+
# a dependency of application charts to inject those utilities and functions into the rendering
12+
# pipeline. Library charts do not define any templates and therefore cannot be deployed.
13+
type: application
14+
15+
# This is the chart version. This version number should be incremented each time you make changes
16+
# to the chart and its templates, including the app version.
17+
# Versions are expected to follow Semantic Versioning (https://semver.org/)
18+
version: 0.1.0
19+
20+
# This is the version number of the application being deployed. This version number should be
21+
# incremented each time you make changes to the application. Versions are not expected to
22+
# follow Semantic Versioning. They should reflect the version the application is using.
23+
# It is recommended to use it with quotes.
24+
appVersion: "1.16.0"
25+
26+
dependencies:
27+
- name: gpu-operator
28+
version: "v25.3.4"
29+
repository: "https://helm.ngc.nvidia.com/nvidia"
30+
condition: gpu-operator.enabled
Lines changed: 314 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,314 @@
1+
version: v1
2+
mig-configs:
3+
# Profile 1
4+
all-disabled:
5+
- devices: all
6+
mig-enabled: false
7+
8+
# Profile 19
9+
all-1g.10gb:
10+
- devices: all
11+
mig-enabled: true
12+
mig-devices:
13+
"1g.10gb": 7
14+
15+
all-1g.5gb:
16+
- devices: all
17+
mig-enabled: true
18+
mig-devices:
19+
"1g.5gb": 7
20+
21+
all-1g.18gb:
22+
- devices: all
23+
mig-enabled: true
24+
mig-devices:
25+
"1g.18gb": 7
26+
27+
# Profile 5
28+
all-3g.40gb:
29+
- devices: all
30+
mig-enabled: true
31+
mig-devices:
32+
"3g.40gb": 2
33+
34+
all-3g.20gb:
35+
- devices: all
36+
mig-enabled: true
37+
mig-devices:
38+
"3g.20gb": 2
39+
40+
all-3g.71gb:
41+
- devices: all
42+
mig-enabled: true
43+
mig-devices:
44+
"3g.71gb": 2
45+
46+
# Profile 1
47+
all-7g.80gb:
48+
- devices: all
49+
mig-enabled: true
50+
mig-devices:
51+
"7g.80gb": 1
52+
53+
all-7g.40gb:
54+
- devices: all
55+
mig-enabled: true
56+
mig-devices:
57+
"7g.40gb": 1
58+
59+
all-7g.141gb:
60+
- devices: all
61+
mig-enabled: true
62+
mig-devices:
63+
"7g.141gb": 1
64+
65+
# Profile 2
66+
mixed-1-3g.40gb-1-4g.40gb:
67+
- devices: all
68+
mig-enabled: true
69+
mig-devices:
70+
"3g.40gb": 1
71+
"4g.40gb": 1
72+
73+
mixed-1-3g.20gb-1-4g.20gb:
74+
- devices: all
75+
mig-enabled: true
76+
mig-devices:
77+
"3g.20gb": 1
78+
"4g.20gb": 1
79+
80+
mixed-1-3g.71gb-1-4g.71gb:
81+
- devices: all
82+
mig-enabled: true
83+
mig-devices:
84+
"3g.71gb": 1
85+
"4g.71gb": 1
86+
87+
# Profile 3
88+
mixed-1-1g.10gb-1-2g.20gb-1-4g.40gb:
89+
- devices: all
90+
mig-enabled: true
91+
mig-devices:
92+
"1g.10gb": 1
93+
"2g.20gb": 1
94+
"4g.40gb": 1
95+
96+
mixed-1-1g.5gb-1-2g.10gb-1-4g.20gb:
97+
- devices: all
98+
mig-enabled: true
99+
mig-devices:
100+
"1g.5gb": 1
101+
"2g.10gb": 1
102+
"4g.20gb": 1
103+
104+
mixed-1-1g.18gb-1-2g.35gb-1-4g.71gb:
105+
- devices: all
106+
mig-enabled: true
107+
mig-devices:
108+
"1g.18gb": 1
109+
"2g.35gb": 1
110+
"4g.71gb": 1
111+
112+
# Profile 4
113+
mixed-3-1g.10gb-1-4g.40gb:
114+
- devices: all
115+
mig-enabled: true
116+
mig-devices:
117+
"1g.10gb": 3
118+
"4g.40gb": 1
119+
120+
mixed-3-1g.5gb-1-4g.20gb:
121+
- devices: all
122+
mig-enabled: true
123+
mig-devices:
124+
"1g.5gb": 3
125+
"4g.20gb": 1
126+
127+
mixed-3-1g.18gb-1-4g.71gb:
128+
- devices: all
129+
mig-enabled: true
130+
mig-devices:
131+
"1g.18gb": 3
132+
"4g.71gb": 1
133+
134+
# Profile 6
135+
mixed-1g.10gb-1-2g.20gb-1-3g.40gb:
136+
- devices: all
137+
mig-enabled: true
138+
mig-devices:
139+
"1g.10gb": 1
140+
"2g.20gb": 1
141+
"3g.40gb": 1
142+
143+
mixed-1g.5gb-1-2g.10gb-1-3g.20gb:
144+
- devices: all
145+
mig-enabled: true
146+
mig-devices:
147+
"1g.5gb": 1
148+
"2g.10gb": 1
149+
"3g.20gb": 1
150+
151+
mixed-1g.18gb-1-2g.35gb-1-3g.71gb:
152+
- devices: all
153+
mig-enabled: true
154+
mig-devices:
155+
"1g.18gb": 1
156+
"2g.35gb": 1
157+
"3g.71gb": 1
158+
159+
# Profile 7
160+
mixed-3-1g.10gb-1-3g.40gb:
161+
- devices: all
162+
mig-enabled: true
163+
mig-devices:
164+
"1g.10gb": 3
165+
"3g.40gb": 1
166+
167+
mixed-3-1g.5gb-1-3g.20gb:
168+
- devices: all
169+
mig-enabled: true
170+
mig-devices:
171+
"1g.5gb": 3
172+
"3g.20gb": 1
173+
174+
mixed-3-1g.18gb-1-3g.70gb:
175+
- devices: all
176+
mig-enabled: true
177+
mig-devices:
178+
"1g.18gb": 3
179+
"3g.71gb": 1
180+
181+
# Profile 8
182+
mixed-2-2g.20gb-1-3g.40gb:
183+
- devices: all
184+
mig-enabled: true
185+
mig-devices:
186+
"2g.20gb": 2
187+
"3g.40gb": 1
188+
189+
mixed-2-2g.10gb-1-3g.20gb:
190+
- devices: all
191+
mig-enabled: true
192+
mig-devices:
193+
"2g.10gb": 2
194+
"3g.20gb": 1
195+
196+
mixed-2-2g.35gb-1-3g.71gb:
197+
- devices: all
198+
mig-enabled: true
199+
mig-devices:
200+
"2g.35gb": 2
201+
"3g.71gb": 1
202+
203+
# Profile 9, 10
204+
mixed-2-1g.10gb-1-2g.20gb-1-3g.40gb:
205+
- devices: all
206+
mig-enabled: true
207+
mig-devices:
208+
"1g.10gb": 2
209+
"2g.20gb": 1
210+
"3g.40gb": 1
211+
212+
mixed-2-1g.5gb-1-2g.20gb-1-3g.20gb:
213+
- devices: all
214+
mig-enabled: true
215+
mig-devices:
216+
"1g.5gb": 2
217+
"2g.10gb": 1
218+
"3g.20gb": 1
219+
220+
mixed-2-1g.18gb-1-2g.47gb-1-3g.71gb:
221+
- devices: all
222+
mig-enabled: true
223+
mig-devices:
224+
"1g.18gb": 2
225+
"2g.47gb": 1
226+
"3g.71gb": 1
227+
228+
# Profile 11
229+
mixed-4-1g.10gb-1-3g.40gb:
230+
- devices: all
231+
mig-enabled: true
232+
mig-devices:
233+
"1g.10gb": 4
234+
"3g.40gb": 1
235+
236+
mixed-4-1g.5gb-1-3g.20gb:
237+
- devices: all
238+
mig-enabled: true
239+
mig-devices:
240+
"1g.5gb": 4
241+
"3g.20gb": 1
242+
243+
mixed-4-1g.18gb-1-3g.71gb:
244+
- devices: all
245+
mig-enabled: true
246+
mig-devices:
247+
"1g.18gb": 4
248+
"3g.71gb": 1
249+
250+
# Profile 12
251+
mixed-1-1g.10gb-3-2g.20gb:
252+
- devices: all
253+
mig-enabled: true
254+
mig-devices:
255+
"1g.10gb": 1
256+
"2g.20gb": 3
257+
258+
mixed-1-1g.5gb-3-2g.10gb:
259+
- devices: all
260+
mig-enabled: true
261+
mig-devices:
262+
"1g.5gb": 1
263+
"2g.10gb": 3
264+
265+
mixed-1-1g.18gb-3-2g.35gb:
266+
- devices: all
267+
mig-enabled: true
268+
mig-devices:
269+
"1g.18gb": 1
270+
"2g.35gb": 3
271+
272+
# Profile 13, 14
273+
mixed-3-1g.10gb-2-2g.20gb:
274+
- devices: all
275+
mig-enabled: true
276+
mig-devices:
277+
"1g.10gb": 3
278+
"2g.20gb": 2
279+
280+
mixed-3-1g.5gb-2-2g.10gb:
281+
- devices: all
282+
mig-enabled: true
283+
mig-devices:
284+
"1g.5gb": 3
285+
"2g.10gb": 2
286+
287+
mixed-3-1g.18gb-2-2g.35gb:
288+
- devices: all
289+
mig-enabled: true
290+
mig-devices:
291+
"1g.18gb": 3
292+
"2g.35gb": 2
293+
294+
# Profile 15, 16, 17, 18, 19
295+
mixed-5-1g.10gb-1-2g.20gb:
296+
- devices: all
297+
mig-enabled: true
298+
mig-devices:
299+
"1g.10gb": 5
300+
"2g.20gb": 1
301+
302+
mixed-5-1g.5gb-1-2g.10gb:
303+
- devices: all
304+
mig-enabled: true
305+
mig-devices:
306+
"1g.5gb": 5
307+
"2g.10gb": 1
308+
309+
mixed-5-1g.18gb-1-2g.35gb:
310+
- devices: all
311+
mig-enabled: true
312+
mig-devices:
313+
"1g.18gb": 5
314+
"2g.35gb": 1
Lines changed: 13 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,13 @@
1+
gpu-operator:
2+
operator:
3+
repository: "248189928082.dkr.ecr.ap-northeast-1.amazonaws.com"
4+
toolkit:
5+
repository: "248189928082.dkr.ecr.ap-northeast-1.amazonaws.com/mirror-k8s"
6+
devicePlugin:
7+
repository: "248189928082.dkr.ecr.ap-northeast-1.amazonaws.com"
8+
gfd:
9+
repository: "248189928082.dkr.ecr.ap-northeast-1.amazonaws.com"
10+
migManager:
11+
repository: "248189928082.dkr.ecr.ap-northeast-1.amazonaws.com/mirror-cloud-native"
12+
validator:
13+
repository: "248189928082.dkr.ecr.ap-northeast-1.amazonaws.com/mirror-cloud-native"
Lines changed: 13 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,13 @@
1+
gpu-operator:
2+
operator:
3+
repository: "183295424886.dkr.ecr.ap-south-1.amazonaws.com"
4+
toolkit:
5+
repository: "183295424886.dkr.ecr.ap-south-1.amazonaws.com/mirror-k8s"
6+
devicePlugin:
7+
repository: "183295424886.dkr.ecr.ap-south-1.amazonaws.com"
8+
gfd:
9+
repository: "183295424886.dkr.ecr.ap-south-1.amazonaws.com"
10+
migManager:
11+
repository: "183295424886.dkr.ecr.ap-south-1.amazonaws.com/mirror-cloud-native"
12+
validator:
13+
repository: "183295424886.dkr.ecr.ap-south-1.amazonaws.com/mirror-cloud-native"

0 commit comments

Comments
 (0)