NVIDIA RTX 4000 SFF Ada vs 2x Radeon Pro Vega II Duo

NVIDIA RTX 4000 SFF Ada
2x Radeon Pro Vega II Duo $11,200
6144 Shaders
20GB GDDR6
1560MHz
4x 4096 Shaders
128GB (4x 32GB) HBM2
1720MHz
Peak AI Performance
613.42 TOPS
INT4 Tensor Sparse
Peak AI Performance
112.72 TFLOPS
FP16
FP32
19.17 TFLOPS
FP32
56.36 TFLOPS
FP16
19.17 TFLOPS
FP16
112.72 TFLOPS
Form Factor
PCIe Card
2.0-Slots
Form Factor
MPX
2x 4.0-Slots
TDP
70W
TDP
950W
-
-
-
-
-
-
-
-
-
-
GB6 OpenCL N/A
0%
GB6 Metal N/A
0%
GB6 Metal N/A
0%
GB6 Vulkan N/A
0%
GB5 OpenCL N/A
0%
GB5 OpenCL N/A
0%
GB5 CUDA N/A
0%
GB5 CUDA N/A
0%
GB5 Metal N/A
0%
GB5 Metal N/A
0%
GB5 Vulkan N/A
0%
GB5 Vulkan N/A
0%
OCT 2020.1 N/A
0%
OCT 2020.1 N/A
0%
OCT Metal N/A
0%
OCT Metal N/A
0%
Peak AI Performance
613.42 TOPS
INT4 Tensor Sparse
Peak AI Performance
112.72 TFLOPS
FP16
-
-
-
-
-
-
FP8
-
153.35 TFLOPS Tensor (FP16 Accumulate)
306.71 TFLOPS Tensor (FP16 Accumulate) Sparse
153.35 TFLOPS Tensor (FP32 Accumulate)
306.71 TFLOPS Tensor (FP32 Accumulate) Sparse
-
-
-
-
-
-
FP16
19.17 TFLOPS
76.68 TFLOPS Tensor (FP16 Accumulate)
153.35 TFLOPS Tensor (FP16 Accumulate) Sparse
76.68 TFLOPS Tensor (FP32 Accumulate)
153.35 TFLOPS Tensor (FP32 Accumulate) Sparse
FP16
112.72 TFLOPS
-
-
-
-
FP32
19.17 TFLOPS
-
-
FP32
56.36 TFLOPS
-
-
FP64
300 GFLOPS
-
FP64
14.09 TFLOPS
-
BF16
19.17 TFLOPS
76.68 TFLOPS Tensor
153.35 TFLOPS Tensor Sparse
-
-
-
-
TF32
38.34 TFLOPS Tensor
76.68 TFLOPS Tensor Sparse
-
-
-
INT4
306.71 TOPS Tensor
613.42 TOPS Tensor Sparse
-
-
-
INT8
-
153.35 TOPS Tensor
306.71 TOPS Tensor Sparse
-
-
-
-
INT32
9.59 TOPS
-
-
Ray Tracing
44.3 TOPS
-
-
Pixel Fillrate
124.8 GPixel/s
Pixel Fillrate
110.08 GPixel/s
-
-
-
-
Texture Fillrate
299.52 GTexel/s
Texture Fillrate
440.32 GTexel/s
Manufacturer
NVIDIA
Manufacturer
Apple
Chip Designer
NVIDIA
Chip Designer
AMD
Architecture
Ada Lovelace
Architecture
GCN 5
Family
RTX Ada
Family
Radeon Pro Vega
Codename
NV184
AD104
-
-
Codename
Moonshot
Vega 20
Variant
Vega 20 XTXA
Market Segment
Workstation
Market Segment
Workstation
Release Date
3/21/2023
Release Date
12/1/2019
Foundry
TSMC
-
Foundry
TSMC
-
Fabrication Node
4N
-
Fabrication Node
N7
-
Die Size
295 mm²
-
Die Size
4x 331 mm²
-
Transistor Count
35.8 Billion
-
Transistor Count
4x 13.2 Billion
-
Transistor Density
121.56M/mm²
-
Transistor Density
39.97M/mm²
-
Form
PCIe Card
Form
MPX
Shading Units
6144 Shaders
-
Shading Units
4x 4096 Shaders
-
Texture Mapping Units
192 TMUs
Texture Mapping Units
4x 256 TMUs
Render Output Units
80 ROPs
Render Output Units
4x 64 ROPs
Tensor Cores
192 T-Cores
-
-
Ray-Tracing Cores
48 RT-Cores
-
-
Streaming Multiprocessors
48 SMs
-
-
-
-
Compute Units
4x 64 CUs
-
-
-
-
-
-
-
-
-
-
720MHz Base
1560MHz
-
-
-
1720MHz
-
-
-
-
L1
64KB/SM Tex
128KB/SM
-
-
L1
-
-
16KB/CU
-
L2
48MB Shared
L2
4MB Shared
-
-
-
-
-
-
20GB
GDDR6
ECC
128GB (4x 32GB)
HBM2
-
Bus Width
160Bit
Bus Width
4096Bit
Clock
1750MHz
Transfer Rate
14GT/s
Bandwidth
280GB/s
Clock
1000MHz
Transfer Rate
2GT/s
Bandwidth
1024GB/s
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
TDP
70W
TDP
950W
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
4x mini-DisplayPort 1.4
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
2x HDMI 2.0
-
-
8x Thunderbolt 3 (40Gbps)
Max Resolution
7680x4320
Max Resolution
7680x4320
Max Resolution Refresh Rate
120Hz
Max Resolution Refresh Rate
60Hz
Variable Refresh Rate
G-Sync
FreeSync
-
Variable Refresh Rate
-
FreeSync
-
Display Stream Compression (DSC)
Supported
Display Stream Compression (DSC)
Not Supported
Multi Monitor Support
4
Multi Monitor Support
3
Content Protection
HDCP 2.3
Content Protection
HDCP 2.2
Model
2x NVENC 8
Model
VCE 4.1
Codec
-
-
-
-
-
-
-
-
AVC (H.264)
HEVC (H.265)
-
AV1
-
-
Codec
-
-
-
-
-
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
Model
2x NVDEC 5
Model
UVD 7.2
Codec
MPEG-1
MPEG-2
MPEG-4
-
VC-1
VP8
VP9
-
AVC (H.264)
HEVC (H.265)
-
AV1
-
-
Codec
MPEG-1
MPEG-2
MPEG-4
JPEG
VC-1
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
Direct X
12
Direct 3D
12_3
Direct X
12
Direct 3D
12_1
OpenGL
4.6
OpenCL
3.0
Vulkan
1.3
OpenGL
4.6
OpenCL
2.1
Vulkan
1.3
Shader Model
6.7
CUDA
8.9
-
-
PureVideo HD
VP12
VDPAU
Feature Set L
Shader Model
6.7
-
-
GFX
9
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Slots Required
2.0
PCIe Version
4.0
PCIe Lanes
16
Slots Required
4.0
PCIe Version
3.0
PCIe Lanes
16
-
-
-
-
Multi GPU Support
Supported
Type
CrossFire XDMA
Height
69 mm (2.72 in)
Width
168 mm (6.61 in)
Depth
40 mm (1.57 in)
-
-
-
-
-
-
Change Comparison