AMD Radeon Instinct MI50 vs NVIDIA V100 FHHL

AMD Radeon Instinct MI50
NVIDIA V100 FHHL
3840 Shaders
16GB HBM2
1725MHz
5120 Shaders
16GB HBM2
1290MHz
Peak AI Performance
26.5 TFLOPS
FP16
Peak AI Performance
105.68 TFLOPS
FP16 Tensor (FP16 Accumulate)
FP32
13.25 TFLOPS
FP32
13.21 TFLOPS
FP16
26.5 TFLOPS
FP16
26.42 TFLOPS
Form Factor
PCIe Card
2.0-Slots
Form Factor
PCIe Card
1.0-Slots
TDP
300W
TDP
250W
Power Connectors
2x 8-Pin
Power Connectors
1x 8-Pin

Peak AI Performance

  • 75% slower vs V100 FHHL
  • 3.99x faster vs Radeon Instinct MI50
Radeon Instinct MI50 - 26.5 TFLOPS FP16
x1
V100 FHHL - 105.68 TFLOPS FP16 Tensor (FP16 Accumulate)
x3.99

FP32

  • Same Performance vs V100 FHHL
  • Same Performance vs Radeon Instinct MI50
Radeon Instinct MI50 - 13.25 TFLOPS FP32
x1
V100 FHHL - 13.21 TFLOPS FP32
x1

FP16

  • Same Performance vs V100 FHHL
  • Same Performance vs Radeon Instinct MI50
Radeon Instinct MI50 - 26.5 TFLOPS FP16
x1
V100 FHHL - 26.42 TFLOPS FP16
x1
  • 23% faster vs V100 FHHL
  • 19% slower vs Radeon Instinct MI50
Radeon Instinct MI50 - 1024GB/s
x1.23
V100 FHHL - 829.4GB/s
x1
  • 20% higher vs V100 FHHL
  • 17% lower vs Radeon Instinct MI50
Radeon Instinct MI50 - 300W
x1.2
V100 FHHL - 250W
x1
Manufacturer
AMD
Manufacturer
NVIDIA
Chip Designer
AMD
Chip Designer
NVIDIA
Architecture
GCN 5
Architecture
Volta
Family
Instinct
Family
Server
Codename
Moonshot
Vega 20
Variant
Vega 20 XL GL
Codename
NV140
GV100
-
-
Market Segment
Server
Market Segment
Server
Release Date
11/18/2018
Release Date
3/27/2018
Foundry
TSMC
Foundry
TSMC
Fabrication Node
N7
Fabrication Node
12FFN
Die Size
331 mm²
Die Size
815 mm²
Transistor Count
13.2 Billion
Transistor Count
21.1 Billion
Transistor Density
39.97M/mm²
Transistor Density
25.89M/mm²
Form
PCIe Card
Form
PCIe Card
Shading Units
3840 Shaders
Shading Units
5120 Shaders
Texture Mapping Units
240 TMUs
Texture Mapping Units
320 TMUs
Render Output Units
64 ROPs
Render Output Units
128 ROPs
-
-
Tensor Cores
640 T-Cores
-
-
Streaming Multiprocessors
80 SMs
Compute Units
60 CUs
-
-
-
-
Graphics Processing Clusters
6 GPCs
1450MHz Base
1725MHz
937MHz Base
1290MHz
Peak AI Performance
26.5 TFLOPS
FP16
Peak AI Performance
105.68 TFLOPS
FP16 Tensor (FP16 Accumulate)
FP16
26.5 TFLOPS
-
FP16
26.42 TFLOPS
105.68 TFLOPS Tensor (FP16 Accumulate)
FP32
13.25 TFLOPS
FP32
13.21 TFLOPS
FP64
6.62 TFLOPS
FP64
6.61 TFLOPS
-
-
INT8
52.84 TOPS
-
-
INT32
13.21 TOPS
Pixel Fillrate
110.4 GPixel/s
Pixel Fillrate
165.12 GPixel/s
Texture Fillrate
414 GTexel/s
Texture Fillrate
412.8 GTexel/s
L1
-
16KB/CU
L1
128KB/SM
-
L2
4MB Shared
L2
6MB Shared
16GB
HBM2
ECC
16GB
HBM2
-
Bus Width
4096Bit
Bus Width
4096Bit
Clock
1000MHz
Transfer Rate
2GT/s
Bandwidth
1024GB/s
Clock
810MHz
Transfer Rate
1.6GT/s
Bandwidth
829.4GB/s
TDP
300W
TDP
250W
1x mini-DisplayPort 1.4
-
Max Resolution
7680x4320
Max Resolution
Unknown
Max Resolution Refresh Rate
60Hz
Max Resolution Refresh Rate
-
Variable Refresh Rate
-
FreeSync
Variable Refresh Rate
G-Sync
FreeSync
Display Stream Compression (DSC)
Not Supported
Display Stream Compression (DSC)
Not Supported
Multi Monitor Support
3
Multi Monitor Support
Unknown
Content Protection
HDCP 2.2
-
-
Model
VCE 4.1
Model
3x NVENC 5
Codec
AVC (H.264)
HEVC (H.265)
Codec
AVC (H.264)
HEVC (H.265)
Model
UVD 7.2
Model
NVDEC 3
Codec
MPEG-1
MPEG-2
MPEG-4
JPEG
VC-1
-
AVC (H.264)
HEVC (H.265)
Codec
MPEG-1
MPEG-2
MPEG-4
-
VC-1
VP9
AVC (H.264)
HEVC (H.265)
Direct X
12
Direct 3D
12_1
Direct X
12
Direct 3D
12_1
OpenGL
4.6
OpenCL
2.1
Vulkan
1.3
OpenGL
4.6
OpenCL
3.0
-
-
Shader Model
6.7
-
-
GFX
9
-
-
-
-
Shader Model
6.7
CUDA
7.0
-
-
PureVideo HD
VP9
VDPAU
Feature Set I
Power Connectors
2x 8-Pin
Power Connectors
1x 8-Pin
Slots Required
2.0
PCIe Version
3.0
PCIe Lanes
16
Slots Required
1.0
PCIe Version
3.0
PCIe Lanes
16
Multi GPU Support
Supported
Type
CrossFire XDMA
Multi GPU Support
Supported
Type
NVLink
Height
111 mm (4.37 in)
Width
267 mm (10.51 in)
Depth
37 mm (1.46 in)
Height
111 mm (4.37 in)
Width
168 mm (6.61 in)
Depth
20 mm (0.79 in)
Change Comparison