NVIDIA GV100 Graphics Processing Unit (GPU)

Overview
Manufacturer
NVIDIA
Series
Tesla Volta
Release Date
May 10th, 2017
Graphics Processing Unit
GPU Model
GV100
Architecture
Volta
Fabrication Process
12 nm FinFET
Die Size
815 mm2
Transistors Count
21.1B
Transistors Density
25.9M TRAN/mm2
CUDAs
5120
SM
80
GPCs
6
TMUs
320
Clocks
Base Clock
1200 MHz
Boost Clock
1455 MHz
Memory Clock
880 MHz
Effective Memory Clock
1760 Mbps
Memory Configuration
Memory Size
16384 MB
Memory Type
HBM2
Memory Bus Width
4096-bit
Memory Bandwidth
901.1 GB/s
Physical
Interface
PCI-Express 3.0 x16
Power Connectors
4 × 6-pin, 4 × 8-pin
Thermal Design Power
1500 W
Display Outputs
DisplayPort
3 ×
API Support
DirectX
12.0
Mantle
-
Vulkan
1.0
OpenGL
4.5
OpenCL
2.1
Shader Model
5.0

Performance
Texture Fillrate
384 GTexel/s
SPFP Performance
12.3 TFLOPS
Performance per W
8.2 GFLOPS/W
Performance per mm2
15.1 GFLOPS/mm2


 ModelGPUClocksMemory
NVIDIA DGX Station GV100 1200 / 880 MHz 16 GB HBM2
NVIDIA DGX-1V GV100 1200 / 880 MHz 16 GB HBM2
NVIDIA DGX-2 GV100 1200 / 880 MHz 32 GB HBM2
NVIDIA HGX-1 GV100 1200 / 880 MHz 16 GB HBM2
NVIDIA Quadro GV100 GV100 1450 / 850 MHz 32 GB HBM2
NVIDIA Tesla V100 FHHL PCIe GV100 1200 / 880 MHz 16 GB HBM2
NVIDIA Tesla V100 PCIe GV100 1200 / 880 MHz 16 GB HBM2
NVIDIA Tesla V100 PCIe 32GB GV100 1200 / 880 MHz 32 GB HBM2
NVIDIA Tesla V100 SMX2 GV100 1200 / 880 MHz 16 GB HBM2
NVIDIA Tesla V100 SMX2 32GB GV100 1200 / 880 MHz 32 GB HBM2
NVIDIA TITAN V GV100-400 1200 / 850 MHz 12 GB HBM2
NVIDIA TITAN V CEO Edition GV100-400 1200 / 850 MHz 32 GB HBM2