NVIDIA vGPU 제품별 Type 과 사양 정리
NVIDIA 의 GRID 제품은 과거 GRID K1, K2 로 나누어진다. GRID K1, K2 의 경우 별도의 NVIDIA vGPU 라이선스 없이 사용이 가능했다.
근래 M, P, V 시리즈의 경우 Tesla 코드명으로 다 통합되었고, GRID vGPU 와 GPGPU 용도 중 하나를 선택해서 사용 가능하다.
단, GRID vGPU 용도로 사용할 경우에는 NVIDIA vGPU 관련 라이선스를 별도로 구매해야 한다.
아래는 M, P, V 제품별 vGPU 규격 정보와 사양 정보이다.
제품 규격명은 <GPU제품명>-<allocated_gpu_memory>_<Q|B|A> 이다.
마지막의 Q, B, A 의 의미는
Q : 무거운 3D 그래픽 전용 작업, 파워유저
B : 가벼운 3D 그래픽 작업, 파워유저
A : XenAPP, RDHS 형태의 원격 접속 세션에서 사용가능한 vGPU
라고 보면 된다. 즉, 매우 무거운 3D 작업을 위해서는 Q 이니셜이 붙는 형태로 vGPU 를 구성해야 한다.
NVIDIA vGPUs with less than 1 Gbyte of frame buffer support only 1 virtual display head on a Windows 10 guest OS.
1.4.1.1. Tesla M60 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
M60-8Q | Designer | 8192 | 4 | 4096×2160 | 1 | 2 | Quadro vDWS |
M60-4Q | Designer | 4096 | 4 | 4096×2160 | 2 | 4 | Quadro vDWS |
M60-2Q | Designer | 2048 | 4 | 4096×2160 | 4 | 8 | Quadro vDWS |
M60-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 8 | 16 | Quadro vDWS |
M60-0Q | Power User, Designer | 512 | 2 | 2560×1600 | 16 | 32 | Quadro vDWS |
M60-2B | Power User | 2048 | 2 | 4096×2160 | 4 | 8 | GRID Virtual PC or Quadro vDWS |
M60-1B | Power User | 1024 | 4 | 2560×1600 | 8 | 16 | GRID Virtual PC or Quadro vDWS |
M60-0B | Power User | 512 | 2 | 2560×1600 | 16 | 32 | GRID Virtual PC or Quadro vDWS |
M60-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 1 | 2 | GRID Virtual Application |
M60-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 2 | 4 | GRID Virtual Application |
M60-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 4 | 8 | GRID Virtual Application |
M60-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 8 | 16 | GRID Virtual Application |
1.4.1.2. Tesla M10 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
M10-8Q | Designer | 8192 | 4 | 4096×2160 | 1 | 4 | Quadro vDWS |
M10-4Q | Designer | 4096 | 4 | 4096×2160 | 2 | 8 | Quadro vDWS |
M10-2Q | Designer | 2048 | 4 | 4096×2160 | 4 | 16 | Quadro vDWS |
M10-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 8 | 32 | Quadro vDWS |
M10-0Q | Power User, Designer | 512 | 2 | 2560×1600 | 16 | 64 | Quadro vDWS |
M10-2B | Power User | 2048 | 2 | 4096×2160 | 4 | 16 | GRID Virtual PC or Quadro vDWS |
M10-1B | Power User | 1024 | 4 | 2560×1600 | 8 | 32 | GRID Virtual PC or Quadro vDWS |
M10-0B | Power User | 512 | 2 | 2560×1600 | 16 | 64 | GRID Virtual PC or Quadro vDWS |
M10-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 1 | 4 | GRID Virtual Application |
M10-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 2 | 8 | GRID Virtual Application |
M10-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 4 | 16 | GRID Virtual Application |
M10-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 8 | 32 | GRID Virtual Application |
1.4.1.3. Tesla M6 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
M6-8Q | Designer | 8192 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
M6-4Q | Designer | 4096 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
M6-2Q | Designer | 2048 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
M6-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 8 | 8 | Quadro vDWS |
M6-0Q | Power User, Designer | 512 | 2 | 2560×1600 | 16 | 16 | Quadro vDWS |
M6-2B | Power User | 2048 | 2 | 4096×2160 | 4 | 4 | GRID Virtual PC or Quadro vDWS |
M6-1B | Power User | 1024 | 4 | 2560×1600 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
M6-0B | Power User | 512 | 2 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
M6-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
M6-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
M6-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
M6-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
1.4.1.4. Tesla P100 PCIe 12GB Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
P100C-12Q | Designer | 12288 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
P100C-6Q | Designer | 6144 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
P100C-4Q | Designer | 4096 | 4 | 4096×2160 | 3 | 3 | Quadro vDWS |
P100C-2Q | Designer | 2048 | 4 | 4096×2160 | 6 | 6 | Quadro vDWS |
P100C-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 12 | 12 | Quadro vDWS |
P100C-2B | Power User | 2048 | 2 | 4096×2160 | 6 | 6 | GRID Virtual PC or Quadro vDWS |
P100C-1B | Power User | 1024 | 4 | 2560×1600 | 12 | 12 | GRID Virtual PC or Quadro vDWS |
P100C-12A | Virtual Application User | 12288 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
P100C-6A | Virtual Application User | 6144 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
P100C-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 3 | 3 | GRID Virtual Application |
P100C-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 6 | 6 | GRID Virtual Application |
P100C-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 12 | 12 | GRID Virtual Application |
1.4.1.5. Tesla P100 PCIe 16GB Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
P100-16Q | Designer | 16384 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
P100-8Q | Designer | 8192 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
P100-4Q | Designer | 4096 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
P100-2Q | Designer | 2048 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
P100-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 16 | 16 | Quadro vDWS |
P100-2B | Power User | 2048 | 2 | 4096×2160 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
P100-1B | Power User | 1024 | 4 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
P100-16A | Virtual Application User | 16384 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
P100-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
P100-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
P100-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
P100-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 16 | 16 | GRID Virtual Application |
1.4.1.6. Tesla P100 SXM2 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
P100X-16Q | Designer | 16384 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
P100X-8Q | Designer | 8192 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
P100X-4Q | Designer | 4096 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
P100X-2Q | Designer | 2048 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
P100X-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 16 | 16 | Quadro vDWS |
P100X-2B | Power User | 2048 | 2 | 4096×2160 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
P100X-1B | Power User | 1024 | 4 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
P100X-16A | Virtual Application User | 16384 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
P100X-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
P100X-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
P100X-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
P100X-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 16 | 16 | GRID Virtual Application |
1.4.1.7. Tesla P40 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
P40-24Q | Designer | 24576 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
P40-12Q | Designer | 12288 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
P40-8Q | Designer | 8192 | 4 | 4096×2160 | 3 | 3 | Quadro vDWS |
P40-6Q | Designer | 6144 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
P40-4Q | Designer | 4096 | 4 | 4096×2160 | 6 | 6 | Quadro vDWS |
P40-3Q | Designer | 3072 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
P40-2Q | Designer | 2048 | 4 | 4096×2160 | 12 | 12 | Quadro vDWS |
P40-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 24 | 24 | Quadro vDWS |
P40-2B | Power User | 2048 | 2 | 4096×2160 | 12 | 12 | GRID Virtual PC or Quadro vDWS |
P40-1B | Power User | 1024 | 4 | 2560×1600 | 24 | 24 | GRID Virtual PC or Quadro vDWS |
P40-24A | Virtual Application User | 24576 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
P40-12A | Virtual Application User | 12288 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
P40-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 3 | 3 | GRID Virtual Application |
P40-6A | Virtual Application User | 6144 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
P40-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 6 | 6 | GRID Virtual Application |
P40-3A | Virtual Application User | 3072 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
P40-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 12 | 12 | GRID Virtual Application |
P40-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 24 | 24 | GRID Virtual Application |
1.4.1.8. Tesla P6 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
P6-16Q | Designer | 16384 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
P6-8Q | Designer | 8192 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
P6-4Q | Designer | 4096 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
P6-2Q | Designer | 2048 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
P6-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 16 | 16 | Quadro vDWS |
P6-2B | Power User | 2048 | 2 | 4096×2160 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
P6-1B | Power User | 1024 | 4 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
P6-16A | Virtual Application User | 16384 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
P6-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
P6-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
P6-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
P6-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 16 | 16 | GRID Virtual Application |
1.4.1.9. Tesla P4 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
P4-8Q | Designer | 8192 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
P4-4Q | Designer | 4096 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
P4-2Q | Designer | 2048 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
P4-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 8 | 8 | Quadro vDWS |
P4-2B | Power User | 2048 | 2 | 4096×2160 | 4 | 4 | GRID Virtual PC or Quadro vDWS |
P4-1B | Power User | 1024 | 4 | 2560×1600 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
P4-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
P4-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
P4-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
P4-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
1.4.1.10. Tesla V100 SXM2 Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
V100X-16Q | Designer | 16384 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
V100X-8Q | Designer | 8192 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
V100X-4Q | Designer | 4096 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
V100X-2Q | Designer | 2048 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
V100X-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 16 | 16 | Quadro vDWS |
V100X-2B | Power User | 2048 | 2 | 4096×2160 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
V100X-1B | Power User | 1024 | 4 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
V100X-16A | Virtual Application User | 16384 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
V100X-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
V100X-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
V100X-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
V100X-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 16 | 16 | GRID Virtual Application |
1.4.1.11. Tesla V100 PCIe Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
V100-16Q | Designer | 16384 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
V100-8Q | Designer | 8192 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
V100-4Q | Designer | 4096 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
V100-2Q | Designer | 2048 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
V100-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 16 | 16 | Quadro vDWS |
V100-2B | Power User | 2048 | 2 | 4096×2160 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
V100-1B | Power User | 1024 | 4 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
V100-16A | Virtual Application User | 16384 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
V100-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
V100-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
V100-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
V100-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 16 | 16 | GRID Virtual Application |
Tesla V100 FHHL Virtual GPU Types
Virtual GPU Type | Intended Use Case | Frame Buffer (Mbytes) | Virtual Display Heads | Maximum Resolution per Display Head | Maximum vGPUs per GPU | Maximum vGPUs per Board | Required License Edition |
---|---|---|---|---|---|---|---|
V100L-16Q | Designer | 16384 | 4 | 4096×2160 | 1 | 1 | Quadro vDWS |
V100L-8Q | Designer | 8192 | 4 | 4096×2160 | 2 | 2 | Quadro vDWS |
V100L-4Q | Designer | 4096 | 4 | 4096×2160 | 4 | 4 | Quadro vDWS |
V100L-2Q | Designer | 2048 | 4 | 4096×2160 | 8 | 8 | Quadro vDWS |
V100L-1Q | Power User, Designer | 1024 | 2 | 4096×2160 | 16 | 16 | Quadro vDWS |
V100L-2B | Power User | 2048 | 2 | 4096×2160 | 8 | 8 | GRID Virtual PC or Quadro vDWS |
V100L-1B | Power User | 1024 | 4 | 2560×1600 | 16 | 16 | GRID Virtual PC or Quadro vDWS |
V100L-16A | Virtual Application User | 16384 | 1 | 1280×1024 | 1 | 1 | GRID Virtual Application |
V100L-8A | Virtual Application User | 8192 | 1 | 1280×1024 | 2 | 2 | GRID Virtual Application |
V100L-4A | Virtual Application User | 4096 | 1 | 1280×1024 | 4 | 4 | GRID Virtual Application |
V100L-2A | Virtual Application User | 2048 | 1 | 1280×1024 | 8 | 8 | GRID Virtual Application |
V100L-1A | Virtual Application User | 1024 | 1 | 1280×1024 | 16 | 16 | GRID Virtual Application |
After reading your article, it reminded me of some things about gate io that I studied before. The content is similar to yours, but your thinking is very special, which gave me a different idea. Thank you. But I still have some questions I want to ask you, I will always pay attention. Thanks.
Can you be more specific about the content of your article? After reading it, I still have some doubts. Hope you can help me.