资源介绍
Resource Introduction
如遇问题请投诉反馈
本资源提供适用于 Red Hat Enterprise Linux 7.0 KVM 平台的 NVIDIA vGPU 5.4 完整驱动包，支持 GPU
虚拟化与多虚拟机共享，适合企业级图形加速、AI 计算与高性能虚拟化部署。
Red Hat Enterprise Linux KVM 7.0 NVIDIA vGPU 5.4 完整驱动包，基于 NVIDIA GRID vGPU
5.4 官方版本构建，完整覆盖宿主机、虚拟机及管理组件。该驱动支持在 RHEL 7.0 KVM 环境中对 NVIDIA
专业级及数据中心级显卡进行虚拟化切分，实现多虚拟机同时共享单块 GPU 资源。适用于 VDI、图形工作站虚拟化、CUDA 计算及 AI
推理等场景，具备良好的稳定性与兼容性，是企业级 Linux 虚拟化 GPU 加速的成熟方案。
最新说明
2018-08-03：(5.4)已更新到7-384.155-386.57，本站同步更新。
2018-05-09：(5.3)已更新到7-384.137-386.37，本站同步更新。
2018-01-17：(5.2)已更新到7-384.111-386.09，本站同步更新。
部署教程(简要步骤)
一、宿主机环境准备
安装 Red Hat Enterprise Linux 7.0
启用并配置 KVM / libvirt
BIOS 中开启：
CPU 虚拟化(VT-x / AMD-V)
IOMMU(VT-d / AMD-Vi)
安装匹配内核版本(建议官方 RHEL 7.0 内核)
二、安装 NVIDIA vGPU 宿主驱动
上传并解压 NVIDIA vGPU 5.4 驱动包
安装依赖：
yum install -y kernel-devel kernel-headers gcc make dkms
安装宿主驱动：
./NVIDIA-Linux-x86_64-*-vgpu.run
重启宿主机并确认模块加载：
lsmod | grep nvidia
三、配置 vGPU 类型
查看支持的 vGPU Profile：
nvidia-smi vgpu -q
在虚拟机 XML 中指定 vGPU 类型(libvirt)
启动虚拟机并确认 vGPU 正常分配
四、虚拟机内驱动安装
在虚拟机系统中安装对应的 NVIDIA vGPU Guest Driver
重启虚拟机
使用以下命令验证：
nvidia-smi
注意事项
NVIDIA vGPU 需要 合法 License 授权 才可长期运行
vGPU 5.4 仅支持特定 NVIDIA GPU 型号(如 Tesla M/P 系列)
驱动版本需与宿主机、虚拟机严格匹配
不支持消费级 GeForce 显卡
内核升级可能导致驱动失效，升级前请做好兼容性确认
文件信息
Addeddate：2025-06-22 14:01:03
Category：Driver
Checksum-blake2b：6e87eecf343e35424c4bed0c9c30c9522beab8ad32f8fd1c25486b7403467013e636dbc5b9d6c01576307901af6e65c1506282e430732167389c63ed34b98d16
Checksum-md5：92fd61999741d9084fdbdfd029dcddec
Checksum-sha1：f8bdfb46bc02e9c3b8515f18b611380aac02d911
Checksum-sha256：6142018a2815b59e081fa2c4e029804396ebe99370f097ceefb021da6829d2a1
Checksum-sha512：751aebbb63541f95c69b81365d86c993104e580eabe2c43fcac9ed61bcfcae0dd92708c9e70371ab718d1478680cc30993f00e4bf212d4b89d17b1e176c7e499
Collection_added：drivers
vintagesoftware
softwarehistory
Downloadid：3d442cda-1779-11ea-b98b-023e2e89f3ee
Downloadtype：Image
Identifier： nvgpu_NVIDIA-GRID-RHEL-7-384.155-386.57.zip_ptf3fq6e
Linktype：Current
Name：vGPU Driver
Platformname：Red Hat Enterprise Linux KVM
Platformversion：7.0
Productfamilies：['vGPU']
Productname：vGPU
Releasedate：2018-08-03
Version：5.4
资源预览
RESOURCE PREVIEW
资源预览
NVIDIA-GRID-RHEL-7-384.155-386.57-www.2na.com.zip
[{"title":"384.155-386.57-grid-gpumodeswitch-user-guide.pdf","size":1671214},{"title":"384.155-386.57-grid-license-server-user-guide.pdf","size":3725526},{"title":"384.155-386.57-grid-licensing-user-guide.pdf","size":2096289},{"title":"384.155-386.57-grid-software-quick-start-guide.pdf","size":3606337},{"title":"384.155-386.57-grid-vgpu-release-notes-red-hat-el-kvm.pdf","size":1603714},{"title":"384.155-386.57-grid-vgpu-user-guide.pdf","size":6016168},{"title":"386.57_grid_win10_32bit_international.exe","size":209634680},{"title":"386.57_grid_win10_server2016_64bit_international.exe","size":314990120},{"title":"386.57_grid_win8_win7_32bit_international.exe","size":189903296},{"title":"386.57_grid_win8_win7_server2012R2_server2008R2_64bit_international.exe","size":269235544},{"title":"NVIDIA-Linux-x86_64-384.155-grid.run","size":87577146},{"title":"Red Hat Enterprise Linux KVM 7.0 NVIDIA vGPU 5.4 \u5b8c\u6574\u9a71\u52a8\u5305.url","size":143}]
NVIDIA-GRID-RHEL-7-384.137-386.37-www.2na.com.zip
[{"title":"384.137-386.37-grid-gpumodeswitch-user-guide.pdf","size":1671253},{"title":"384.137-386.37-grid-license-server-release-notes.pdf","size":1686579},{"title":"384.137-386.37-grid-license-server-user-guide.pdf","size":3872155},{"title":"384.137-386.37-grid-licensing-user-guide.pdf","size":2096069},{"title":"384.137-386.37-grid-software-quick-start-guide.pdf","size":3659685},{"title":"384.137-386.37-grid-vgpu-release-notes-red-hat-el-kvm.pdf","size":1601724},{"title":"384.137-386.37-grid-vgpu-user-guide.pdf","size":6920397},{"title":"386.37_grid_win10_32bit_international.exe","size":209233984},{"title":"386.37_grid_win10_server2016_64bit_international.exe","size":313822272},{"title":"386.37_grid_win8_win7_32bit_international.exe","size":189340128},{"title":"386.37_grid_win8_win7_server2012R2_server2008R2_64bit_international.exe","size":269152632},{"title":"NVIDIA-Linux-x86_64-384.137-grid.run","size":84241797},{"title":"Red Hat Enterprise Linux KVM 7.0 NVIDIA vGPU 5.4 \u5b8c\u6574\u9a71\u52a8\u5305.url","size":143}]
NVIDIA-GRID-RHEL-7-384.111-386.09-www.2na.com.zip
[{"title":"384.111-386.09-grid-gpumodeswitch-user-guide.pdf","size":1670743},{"title":"384.111-386.09-grid-license-server-release-notes.pdf","size":1599525},{"title":"384.111-386.09-grid-license-server-user-guide.pdf","size":3825966},{"title":"384.111-386.09-grid-licensing-user-guide.pdf","size":2090375},{"title":"384.111-386.09-grid-software-quick-start-guide.pdf","size":3672206},{"title":"384.111-386.09-grid-vgpu-release-notes-red-hat-el-kvm.pdf","size":1582110},{"title":"386.09_grid_win10_32bit_international.exe","size":208220472},{"title":"386.09_grid_win10_server2016_64bit_international.exe","size":311261352},{"title":"386.09_grid_win8_win7_32bit_international.exe","size":188743848},{"title":"386.09_grid_win8_win7_server2012R2_server2008R2_64bit_international.exe","size":267371344},{"title":"NVIDIA-Linux-x86_64-384.111-grid.run","size":82841207},{"title":"Red Hat Enterprise Linux KVM 7.0 NVIDIA vGPU 5.4 \u5b8c\u6574\u9a71\u52a8\u5305.url","size":143}]
0
·
收藏
·
推广
官方,安全,无毒,稳定,纯净,无广
微信支付
