Intel Gaudi 3 vs AMD Radeon Pro V520

Intel Gaudi 3
AMD Radeon Pro V520
16384 Shaders
128GB HBM2e
438MHz
2304 Shaders
8GB HBM2
1600MHz
Peak AI Performance
1.84 PFLOPS
FP8 Tensor (FP32 Accumulate)
Peak AI Performance
7.37 TFLOPS
FP32
FP32
14.35 TFLOPS
FP32
7.37 TFLOPS
FP16
28.7 TFLOPS
-
-
Form Factor
OAM Module
-
Form Factor
PCIe Card
2.0-Slots
TDP
900W
TDP
225W
-
-
-
-
-
Power Connectors
-
1x 8-Pin
-
-
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
1.84 PFLOPS
FP8 Tensor (FP32 Accumulate)
Peak AI Performance
7.37 TFLOPS
FP32
-
-
-
-
-
-
FP8
-
-
-
1.84 PFLOPS Tensor (FP32 Accumulate)
-
-
-
-
-
-
-
FP16
28.7 TFLOPS
-
-
458.75 TFLOPS Tensor (FP32 Accumulate)
-
-
-
-
-
-
-
FP32
14.35 TFLOPS
229.38 TFLOPS Tensor
-
FP32
7.37 TFLOPS
-
-
-
-
-
FP64
460 GFLOPS
-
BF16
28.7 TFLOPS
1.84 PFLOPS Tensor
-
-
-
-
-
TF32
458.75 TFLOPS Tensor
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Pixel Fillrate
-
Pixel Fillrate
102.4 GPixel/s
-
-
-
-
Texture Fillrate
-
Texture Fillrate
230.4 GTexel/s
Manufacturer
Intel
Manufacturer
AMD
Chip Designer
Intel
Chip Designer
AMD
Architecture
Gaudi 3
Architecture
RDNA 1
Family
Gaudi
Family
Radeon Pro V
Codename
Gaudi 3
Gaudi 3
-
-
Codename
-
Navi 12
Variant
Navi 12 XL GL
Market Segment
Server
Market Segment
Server
Release Date
4/9/2024
Release Date
12/1/2020
Foundry
TSMC
-
Foundry
TSMC
-
Fabrication Node
N5
-
Fabrication Node
N7
-
-
-
-
Die Size
251 mm²
-
-
-
Transistor Count
10.3 Billion
-
-
-
-
Transistor Density
41.04M/mm²
-
Form
OAM Module
Form
PCIe Card
Shading Units
16384 Shaders
-
Shading Units
2304 Shaders
-
Texture Mapping Units
-
Texture Mapping Units
144 TMUs
Render Output Units
-
Render Output Units
64 ROPs
Tensor Cores
64 T-Cores
-
-
-
-
-
-
-
-
-
-
Compute Units
8 CUs
Compute Units
36 CUs
-
-
-
-
-
-
-
-
-
-
-
438MHz
-
-
1000MHz Base
1600MHz
-
-
L0
32KB/WGP
L1
-
-
-
Unknown
L1
-
-
-
128KB/Array
L2
Unknown
L2
4MB Shared
-
-
-
-
-
-
128GB
HBM2e
ECC
8GB
HBM2
-
Bus Width
8192Bit
Bus Width
2048Bit
Clock
1808MHz
Transfer Rate
3.6GT/s
Bandwidth
3702.8GB/s
Clock
1000MHz
Transfer Rate
2GT/s
Bandwidth
512GB/s
-
-
-
-
eSRAM
96MB
12800GB/s
-
-
-
-
-
-
-
-
-
-
-
TDP
900W
TDP
225W
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
No Ports
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
No Ports
Max Resolution
Unknown
Max Resolution
7680x4320
Max Resolution Refresh Rate
-
Max Resolution Refresh Rate
120Hz
Variable Refresh Rate
-
-
-
Variable Refresh Rate
-
FreeSync
-
Display Stream Compression (DSC)
Not Supported
Display Stream Compression (DSC)
Supported
Multi Monitor Support
Unknown
Multi Monitor Support
4
-
-
Content Protection
HDCP 2.3
No Encoders
-
Model
VCN 2.0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Codec
-
-
-
-
-
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
No Decoders
Model
VCN 2.0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Codec
MPEG-1
MPEG-2
MPEG-4
JPEG
VC-1
-
VP9
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
-
-
-
-
Direct X
12
Direct 3D
12_1
-
-
OpenCL
3.0
-
-
OpenGL
4.6
OpenCL
2.1
Vulkan
1.2
-
-
-
-
-
-
-
-
-
-
Shader Model
6.5
-
-
GFX
10.1
-
-
-
-
Not a Card
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Power Connectors
-
-
-
1x 8-Pin
-
-
-
-
-
PCIe Version
5.0
PCIe Lanes
16
Slots Required
2.0
PCIe Version
4.0
PCIe Lanes
16
Multi GPU Support
Supported
Type
RoCE
Multi GPU Support
Supported
Type
CrossFire XDMA
-
-
-
-
-
-
Height
111 mm (4.37 in)
Width
267 mm (10.51 in)
Depth
37 mm (1.46 in)
Change Comparison