NVIDIA CMP 50HX vs Intel Data Center GPU Max Subsystem
GPU Specs Compare
Graphics Card
Release Date
Jun 2021
Jan 2023
Generation
Mining GPUs
Data Center GPU
Type
Desktop
Professional
Bus Interface
PCIe 1.0 x4
PCIe 5.0 x16
Clock Speeds
Base Clock
1350 MHz
900 MHz
Boost Clock
1545 MHz
1600 MHz
Memory Clock
1750 MHz
1565 MHz
Memory
Memory Size
10GB
128GB
Memory Type
GDDR6
HBM2e
Memory Bus
320bit
8192bit
Bandwidth
560.0GB/s
3205GB/s
Render Config
SM Count
56
-
Shading Units
3584
16384
TMUs
192
1024
ROPs
80
0
Tensor Cores
448
1024
RT Cores
56
128
L1 Cache
64 KB (per SM)
64 KB (per EU)
L2 Cache
5 MB
408 MB
Theoretical Performance
Pixel Rate
123.6 GPixel/s
0 MPixel/s
Texture Rate
296.6 GTexel/s
1638 GTexel/s
FP16 (half)
22.15 TFLOPS
52.43 TFLOPS
FP32 (float)
11.07 TFLOPS
52.43 TFLOPS
FP64 (double)
346.1 GFLOPS
52.43 TFLOPS
Graphics Processor
GPU Name
TU102
Ponte Vecchio
GPU Variant
TU102-100-A1
-
Architecture
Turing
Generation 12.5
Foundry
TSMC
Intel
Process Size
12 nm
10 nm
Transistors
18.6 billion
100 billion
Die Size
754 mm²
1280 mm²
Board Design
TDP
250W
2400W
Suggested PSU
600 W
2800 W
Outputs
No outputs
No outputs
Power Connectors
2x 8-pin
1x 16-pin
Graphics Features
DirectX
12 Ultimate (12_2)
12 (12_1)
OpenGL
4.6
4.6
OpenCL
3.0
3.0
Vulkan
1.3
N/A
CUDA
7.5
-
Shader Model
6.8
6.6