Tesla V100 PCIe and Radeon Vega 8 Efficient - video card comparison
Tesla V100 PCIe
Radeon Vega 8 Efficient
Architecture
Volta (2017−2020)
GCN 5.0 (2017−2020)
GPU code name
GV100
Raven
Market segment
Workstation
Desktop
Release date
June 2017
April 2018
Interface
PCIe 3.0 x16
IGP
Supplementary power connectors
2x 8-pin
None
Boost clock speed
1380 MHz
1100 MHz
Core clock speed
1246 MHz
300 MHz
Floating-point performance
14,131 gflops
-
Manufacturing process technology
12 nm
14 nm
Number of transistors
21,100 million
4,940 million
Pipelines / CUDA cores
5120
512
Power consumption (TDP)
250 Watt
35 Watt
Texture fill rate
441.6
35.20
Memory type
HBM2
System Shared
Maximum RAM amount
16 GB
System Shared
Memory bus width
4096 Bit
System Shared
Memory clock speed
1758 MHz
System Shared
Memory bandwidth
900.1 GB/s
-
Display Connectors
No outputs
No outputs
DirectX
12.0
12 (12_1)
CUDA
7.0
-
OpenCL
2.0
2.0
OpenGL
4.6
4.6
Shader Model
5.0
6.4
Vulkan
-
1.2.131