AMD Instinct MI300X vs AMD Radeon Instinct MI50

AMD Instinct MI300X
AMD Radeon Instinct MI50
19456 Shaders
192GB HBM3
2100MHz
3840 Shaders
16GB HBM2
1725MHz
Peak AI Performance
5.23 POPS
INT4 Tensor Sparse
Peak AI Performance
26.5 TFLOPS
FP16
FP32
163.43 TFLOPS
FP32
13.25 TFLOPS
FP16
163.43 TFLOPS
FP16
26.5 TFLOPS
Form Factor
OAM Module
-
Form Factor
PCIe Card
2.0-Slots
TDP
560W
TDP
300W
-
-
-
-
-
Power Connectors
-
2x 8-Pin
-
-
GB6 OpenCL N/A
0%
GB6 Metal N/A
0%
GB6 Metal N/A
0%
GB6 Vulkan 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
5.23 POPS
INT4 Tensor Sparse
Peak AI Performance
26.5 TFLOPS
FP16
-
-
-
-
-
-
FP8
-
2.61 PFLOPS Tensor (FP16 Accumulate)
5.23 PFLOPS Tensor (FP16 Accumulate) Sparse
2.61 PFLOPS Tensor (FP32 Accumulate)
5.23 PFLOPS Tensor (FP32 Accumulate) Sparse
-
-
-
-
-
-
FP16
163.43 TFLOPS
1.31 PFLOPS Tensor (FP16 Accumulate)
2.61 PFLOPS Tensor (FP16 Accumulate) Sparse
1.31 PFLOPS Tensor (FP32 Accumulate)
2.61 PFLOPS Tensor (FP32 Accumulate) Sparse
FP16
26.5 TFLOPS
-
-
-
-
FP32
163.43 TFLOPS
163.43 TFLOPS Tensor
326.86 TFLOPS Tensor Sparse
FP32
13.25 TFLOPS
-
-
FP64
81.72 TFLOPS
163.43 TFLOPS Tensor
FP64
6.62 TFLOPS
-
BF16
-
1.31 PFLOPS Tensor
2.61 PFLOPS Tensor Sparse
-
-
-
-
TF32
653.72 TFLOPS Tensor
1.31 PFLOPS Tensor Sparse
-
-
-
INT4
2.61 POPS Tensor
5.23 POPS Tensor Sparse
-
-
-
INT8
-
2.61 POPS Tensor
5.23 POPS Tensor Sparse
-
-
-
-
-
-
-
-
-
-
-
-
Pixel Fillrate
-
Pixel Fillrate
110.4 GPixel/s
-
-
-
-
Texture Fillrate
2553.6 GTexel/s
Texture Fillrate
414 GTexel/s
Manufacturer
AMD
Manufacturer
AMD
Chip Designer
AMD
Chip Designer
AMD
Architecture
CDNA 3
Architecture
GCN 5
Family
Instinct
Family
Instinct
Codename
Aqua Vanjaram
Aqua Vanjaram XTX
Variant
Aqua Vanjaram XTX
Codename
Moonshot
Vega 20
Variant
Vega 20 XL GL
Market Segment
Server
Market Segment
Server
Release Date
12/6/2023
Release Date
11/18/2018
Foundry
TSMC
TSMC Active Interposer Die
Foundry
TSMC
-
Fabrication Node
N5
N6 Active Interposer Die
Fabrication Node
N7
-
Die Size
4x 370 mm²
-
Die Size
331 mm²
-
Transistor Count
4x 36.5 Billion
-
Transistor Count
13.2 Billion
-
Transistor Density
98.65M/mm²
-
Transistor Density
39.97M/mm²
-
Form
OAM Module
Form
PCIe Card
Shading Units
19456 Shaders
-
Shading Units
3840 Shaders
-
Texture Mapping Units
1216 TMUs
Texture Mapping Units
240 TMUs
Render Output Units
-
Render Output Units
64 ROPs
Tensor Cores
1216 T-Cores
-
-
-
-
-
-
-
-
-
-
Compute Units
304 CUs
Compute Units
60 CUs
-
-
-
-
-
-
-
-
-
-
1000MHz Base
2100MHz
-
-
1450MHz Base
1725MHz
-
-
-
-
L1
-
-
16KB/CU
-
L1
-
-
16KB/CU
-
L2
16MB Shared
L2
4MB Shared
L3
256MB Shared
-
-
-
-
192GB
HBM3
ECC
16GB
HBM2
ECC
Bus Width
8192Bit
Bus Width
4096Bit
Clock
2600MHz
Transfer Rate
5.2GT/s
Bandwidth
5324.8GB/s
Clock
1000MHz
Transfer Rate
2GT/s
Bandwidth
1024GB/s
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
TDP
560W
TDP
300W
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
No Ports
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1x mini-DisplayPort 1.4
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Max Resolution
Unknown
Max Resolution
7680x4320
Max Resolution Refresh Rate
-
Max Resolution Refresh Rate
60Hz
Variable Refresh Rate
-
-
-
Variable Refresh Rate
-
FreeSync
-
Display Stream Compression (DSC)
Not Supported
Display Stream Compression (DSC)
Not Supported
Multi Monitor Support
Unknown
Multi Monitor Support
3
-
-
Content Protection
HDCP 2.2
Model
2x VCN 2.6
Model
VCE 4.1
Codec
-
-
-
-
-
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
Codec
-
-
-
-
-
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
No Decoders
Model
UVD 7.2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Codec
MPEG-1
MPEG-2
MPEG-4
JPEG
VC-1
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
-
-
-
-
Direct X
12
Direct 3D
12_1
-
-
OpenCL
3.0
-
-
OpenGL
4.6
OpenCL
2.1
Vulkan
1.3
-
-
-
-
GFX
9.4
-
-
-
-
Shader Model
6.7
-
-
GFX
9
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Power Connectors
-
-
-
2x 8-Pin
-
-
-
-
-
PCIe Version
5.0
PCIe Lanes
16
Slots Required
2.0
PCIe Version
3.0
PCIe Lanes
16
Multi GPU Support
Supported
Type
Infinity Fabric
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