サーバープロファイル
IBM CloudのVSI for VPC環境における、gx3d-24x120x1a100p
というNVIDIA A100が1台搭載されている環境を利用
環境の確認とNVIDIA Driverのインストール
インストール手順はこちらを参考にしました。
モジュールの最新化
[root@syasuda-ollama ~]# dnf -y update
[root@syasuda-ollama ~]# reboot
環境の確認
[root@syasuda-ollama ~]# cat /etc/redhat-release
Red Hat Enterprise Linux release 9.6 (Plow)
[root@syasuda-ollama ~]# uname -a
Linux syasuda-ollama 5.14.0-570.37.1.el9_6.x86_64 #1 SMP PREEMPT_DYNAMIC Sat Aug 16 01:10:00 EDT 2025 x86_64 x86_64 x86_64 GNU/Linux
[root@syasuda-ollama ~]# hostnamectl
Static hostname: syasuda-ollama
Icon name: computer-vm
Chassis: vm 🖴
Machine ID: 09f2d8a3f371b99ff6e334c58749ec98
Boot ID: 4032088c354a49efb3481bd2e061d564
Virtualization: kvm
Operating System: Red Hat Enterprise Linux 9.6 (Plow)
CPE OS Name: cpe:/o:redhat:enterprise_linux:9::baseos
Kernel: Linux 5.14.0-570.37.1.el9_6.x86_64
Architecture: x86-64
Hardware Vendor: QEMU
Hardware Model: Standard PC _i440FX + PIIX, 1996_
Firmware Version: 1.15.0-1
[root@syasuda-ollama ~]# lspci | grep -e VGA -e NVIDIA
00:02.0 VGA compatible controller: Cirrus Logic GD 5446
04:01.0 3D controller: NVIDIA Corporation GA100 [A100 PCIe 80GB] (rev a1)
現行のrepositoryの確認
[root@syasuda-ollama ~]# yum repolist all
Updating Subscription Management repositories.
This system has release set to 9.6 and it receives updates only for this release.
repo id repo name status
codeready-builder-for-rhel-9-x86_64-eus-rpms Red Hat CodeReady Linux Builder for RHEL 9 x86_64 - Extended Update Support (RPMs) disabled
codeready-builder-for-rhel-9-x86_64-rpms Red Hat CodeReady Linux Builder for RHEL 9 x86_64 (RPMs) disabled
rhel-9-for-x86_64-appstream-eus-rpms Red Hat Enterprise Linux 9 for x86_64 - AppStream - Extended Update Support (RPMs) enabled
rhel-9-for-x86_64-appstream-rpms Red Hat Enterprise Linux 9 for x86_64 - AppStream (RPMs) enabled
rhel-9-for-x86_64-baseos-eus-rpms Red Hat Enterprise Linux 9 for x86_64 - BaseOS - Extended Update Support (RPMs) enabled
rhel-9-for-x86_64-baseos-rpms Red Hat Enterprise Linux 9 for x86_64 - BaseOS (RPMs) enabled
rhel-9-for-x86_64-supplementary-eus-rpms Red Hat Enterprise Linux 9 for x86_64 - Supplementary - Extended Update Support (RPMs) enabled
rhel-9-for-x86_64-supplementary-rpms Red Hat Enterprise Linux 9 for x86_64 - Supplementary (RPMs) enabled
repositoryの追加設定
[root@syasuda-ollama ~]# export distro=rhel9
[root@syasuda-ollama ~]# export arch=x86_64
[root@syasuda-ollama ~]# subscription-manager repos --enable=rhel-9-for-$arch-appstream-rpms
[root@syasuda-ollama ~]# subscription-manager repos --enable=rhel-9-for-$arch-baseos-rpms
[root@syasuda-ollama ~]# subscription-manager repos --enable=codeready-builder-for-rhel-9-$arch-rpms
[root@syasuda-ollama ~]# dnf install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-9.noarch.rpm
[root@syasuda-ollama ~]# dnf config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/$distro/$arch/cuda-$distro.repo
[root@syasuda-ollama ~]# dnf clean expire-cache
[root@syasuda-ollama ~]# yum repolist all
Updating Subscription Management repositories.
This system has release set to 9.6 and it receives updates only for this release.
repo id repo name status
codeready-builder-for-rhel-9-x86_64-eus-rpms Red Hat CodeReady Linux Builder for RHEL 9 x86_64 - Extended Update Support (RPMs) disabled
codeready-builder-for-rhel-9-x86_64-rpms Red Hat CodeReady Linux Builder for RHEL 9 x86_64 (RPMs) enabled
cuda-rhel9-x86_64 cuda-rhel9-x86_64 enabled
epel Extra Packages for Enterprise Linux 9 - x86_64 enabled
epel-cisco-openh264 Extra Packages for Enterprise Linux 9 openh264 (From Cisco) - x86_64 enabled
epel-cisco-openh264-debuginfo Extra Packages for Enterprise Linux 9 openh264 (From Cisco) - x86_64 - Debug disabled
epel-cisco-openh264-source Extra Packages for Enterprise Linux 9 openh264 (From Cisco) - x86_64 - Source disabled
epel-debuginfo Extra Packages for Enterprise Linux 9 - x86_64 - Debug disabled
epel-source Extra Packages for Enterprise Linux 9 - x86_64 - Source disabled
epel-testing Extra Packages for Enterprise Linux 9 - Testing - x86_64 disabled
epel-testing-debuginfo Extra Packages for Enterprise Linux 9 - Testing - x86_64 - Debug disabled
epel-testing-source Extra Packages for Enterprise Linux 9 - Testing - x86_64 - Source disabled
rhel-9-for-x86_64-appstream-eus-rpms Red Hat Enterprise Linux 9 for x86_64 - AppStream - Extended Update Support (RPMs) enabled
rhel-9-for-x86_64-appstream-rpms Red Hat Enterprise Linux 9 for x86_64 - AppStream (RPMs) enabled
rhel-9-for-x86_64-baseos-eus-rpms Red Hat Enterprise Linux 9 for x86_64 - BaseOS - Extended Update Support (RPMs) enabled
rhel-9-for-x86_64-baseos-rpms Red Hat Enterprise Linux 9 for x86_64 - BaseOS (RPMs) enabled
rhel-9-for-x86_64-supplementary-eus-rpms Red Hat Enterprise Linux 9 for x86_64 - Supplementary - Extended Update Support (RPMs) enabled
rhel-9-for-x86_64-supplementary-rpms Red Hat Enterprise Linux 9 for x86_64 - Supplementary (RPMs) enabled
Driverのインストール
[root@syasuda-ollama ~]# dnf -y module install nvidia-driver:latest-dkms
[root@syasuda-ollama ~]# reboot
GPUの確認。NVIDIA A100 80GB PCIeが認識されている。
[root@syasuda-ollama ~]# nvidia-smi
Wed Aug 27 00:05:31 2025
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 580.65.06 Driver Version: 580.65.06 CUDA Version: 13.0 |
+-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA A100 80GB PCIe Off | 00000000:04:01.0 Off | 0 |
| N/A 42C P0 48W / 300W | 0MiB / 81920MiB | 0% Default |
| | | Disabled |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| No running processes found |
+-----------------------------------------------------------------------------------------+
モニタリング目的で、nvtopもインストールしておく
[root@syasuda-ollama ~]# dnf install -y nvtop