diff --git a/docs/ascend910b-support.md b/docs/ascend910b-support.md index 968f9d30d..89420153f 100644 --- a/docs/ascend910b-support.md +++ b/docs/ascend910b-support.md @@ -10,7 +10,7 @@ ## Prerequisites -* Ascend device type: 910B,910B3,910B4,310P +* Ascend device type: 910B,910B2,910B3,910B4,310P * driver version >= 24.1.rc1 * Ascend docker runtime @@ -81,6 +81,27 @@ You can customize the ascend share configuration by following the steps below: - name: vir16 memory: 17476 aiCore: 16 + - chipName: 910B2 + commonWord: Ascend910B2 + resourceName: huawei.com/Ascend910B2 + resourceMemoryName: huawei.com/Ascend910B2-memory + memoryAllocatable: 65536 + memoryCapacity: 65536 + aiCore: 24 + aiCPU: 6 + templates: + - name: vir03_1c_8g + memory: 8192 + aiCore: 3 + aiCPU: 1 + - name: vir06_1c_16g + memory: 16384 + aiCore: 6 + aiCPU: 1 + - name: vir12_3c_32g + memory: 32768 + aiCore: 12 + aiCPU: 3 - chipName: 910B3 commonWord: Ascend910B resourceName: huawei.com/Ascend910B @@ -153,7 +174,63 @@ using the `huawei.com/ascend910` and `huawei.com/ascend910-memory` resource type apiVersion: v1 kind: Pod metadata: - name: gpu-pod + name: ascend910A-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910A: 1 # requesting 1 vGPUs + huawei.com/Ascend910A-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend910B2-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B2: 1 # requesting 1 vGPUs + huawei.com/Ascend910B2-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend910B-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B: 1 # requesting 1 vGPUs + huawei.com/Ascend910B-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend910B4-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B4: 1 # requesting 1 vGPUs + huawei.com/Ascend910B4-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend310P-pod spec: containers: - name: ubuntu-container @@ -161,8 +238,8 @@ spec: command: ["bash", "-c", "sleep 86400"] resources: limits: - huawei.com/Ascend910: 1 # requesting 1 vGPUs - huawei.com/Ascend910-memory: 2000 # requesting 2000m device memory + huawei.com/Ascend310P: 1 # requesting 1 vGPUs + huawei.com/Ascend310P-memory: 2000 # requesting 2000m device memory ``` ## Notes diff --git a/docs/ascend910b-support_cn.md b/docs/ascend910b-support_cn.md index 1fa3cd22d..23f894835 100644 --- a/docs/ascend910b-support_cn.md +++ b/docs/ascend910b-support_cn.md @@ -12,7 +12,7 @@ HAMi 支持复用华为升腾 910B 设备,并为此提供以下几种与 vGPU * Ascend docker runtime * driver version > 24.1.rc1 -* Ascend device type: 910B,910B3,910B4,310P +* Ascend device type: 910B,910B2,910B3,910B4,310P ## 开启 NPU 复用 @@ -81,6 +81,27 @@ HAMi 目前有一个 NPU 内置[虚拟化配置文件](https://github.com/Projec - name: vir16 memory: 17476 aiCore: 16 + - chipName: 910B2 + commonWord: Ascend910B2 + resourceName: huawei.com/Ascend910B2 + resourceMemoryName: huawei.com/Ascend910B2-memory + memoryAllocatable: 65536 + memoryCapacity: 65536 + aiCore: 24 + aiCPU: 6 + templates: + - name: vir03_1c_8g + memory: 8192 + aiCore: 3 + aiCPU: 1 + - name: vir06_1c_16g + memory: 16384 + aiCore: 6 + aiCPU: 1 + - name: vir12_3c_32g + memory: 32768 + aiCore: 12 + aiCPU: 3 - chipName: 910B3 commonWord: Ascend910B resourceName: huawei.com/Ascend910B @@ -153,7 +174,63 @@ HAMi 目前有一个 NPU 内置[虚拟化配置文件](https://github.com/Projec apiVersion: v1 kind: Pod metadata: - name: gpu-pod + name: ascend910A-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910A: 1 # requesting 1 vGPUs + huawei.com/Ascend910A-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend910B2-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B2: 1 # requesting 1 vGPUs + huawei.com/Ascend910B2-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend910B-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B: 1 # requesting 1 vGPUs + huawei.com/Ascend910B-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend910B4-pod +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B4: 1 # requesting 1 vGPUs + huawei.com/Ascend910B4-memory: 2000 # requesting 2000m device memory +--- +apiVersion: v1 +kind: Pod +metadata: + name: ascend310P-pod spec: containers: - name: ubuntu-container @@ -161,8 +238,8 @@ spec: command: ["bash", "-c", "sleep 86400"] resources: limits: - huawei.com/Ascend910: 1 # 请求 1 个 vGPU - huawei.com/Ascend910-memory: 2000 # 请求 2000m 设备内容 + huawei.com/Ascend310P: 1 # requesting 1 vGPUs + huawei.com/Ascend310P-memory: 2000 # requesting 2000m device memory ``` ## 注意事项 diff --git a/examples/ascend/job-310P.yaml b/examples/ascend/job-310P.yaml new file mode 100644 index 000000000..e84431a25 --- /dev/null +++ b/examples/ascend/job-310P.yaml @@ -0,0 +1,13 @@ +apiVersion: v1 +kind: Pod +metadata: + name: ascend310p-job +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend310P: 1 # requesting 1 NPU + huawei.com/Ascend310P-memory: 2000 # requesting 2000m device memory \ No newline at end of file diff --git a/examples/ascend/job-910A.yaml b/examples/ascend/job-910A.yaml new file mode 100644 index 000000000..7f5950173 --- /dev/null +++ b/examples/ascend/job-910A.yaml @@ -0,0 +1,13 @@ +apiVersion: v1 +kind: Pod +metadata: + name: ascend910a-job +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910A: 1 # requesting 1 NPU + huawei.com/Ascend910A-memory: 2000 # requesting 2000m device memory \ No newline at end of file diff --git a/examples/ascend/job-910B2.yaml b/examples/ascend/job-910B2.yaml new file mode 100644 index 000000000..24687bcb1 --- /dev/null +++ b/examples/ascend/job-910B2.yaml @@ -0,0 +1,13 @@ +apiVersion: v1 +kind: Pod +metadata: + name: ascend910b2-job +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B2: 1 # requesting 1 NPU + huawei.com/Ascend910B2-memory: 2000 # requesting 2000m device memory \ No newline at end of file diff --git a/examples/ascend/job-910B3.yaml b/examples/ascend/job-910B3.yaml new file mode 100644 index 000000000..bcdcd9c6a --- /dev/null +++ b/examples/ascend/job-910B3.yaml @@ -0,0 +1,13 @@ +apiVersion: v1 +kind: Pod +metadata: + name: ascend910b-job +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B: 1 # requesting 1 NPU + huawei.com/Ascend910B-memory: 2000 # requesting 2000m device memory \ No newline at end of file diff --git a/examples/ascend/job-910B4.yaml b/examples/ascend/job-910B4.yaml new file mode 100644 index 000000000..09d81634b --- /dev/null +++ b/examples/ascend/job-910B4.yaml @@ -0,0 +1,13 @@ +apiVersion: v1 +kind: Pod +metadata: + name: ascend910b4-job +spec: + containers: + - name: ubuntu-container + image: ascendhub.huawei.com/public-ascendhub/ascend-mindspore:23.0.RC3-centos7 + command: ["bash", "-c", "sleep 86400"] + resources: + limits: + huawei.com/Ascend910B4: 1 # requesting 1 NPU + huawei.com/Ascend910B4-memory: 2000 # requesting 2000m device memory \ No newline at end of file