NVIDIA RTX PRO 4000 Blackwell SFF vs NVIDIA PG506 232
GPU · Specs Compare
Graphics Card
Release Date
Aug 2025
Apr 2021
Generation
Blackwell PRO W
Tesla
Type
Desktop
Desktop
Bus Interface
PCIe 5.0 x8
PCIe 4.0 x16
Clock Speeds
Base Clock
577 MHz
930 MHz
Boost Clock
1432 MHz
1440 MHz
Memory Clock
1125 MHz
1215 MHz
Memory
Memory Size
24GB
24GB
Memory Type
GDDR7
HBM2
Memory Bus
192bit
3072bit
Bandwidth
432.0GB/s
933.1GB/s
Render Config
SM Count
70
56
Shading Units
8960
3584
TMUs
280
224
ROPs
96
96
Tensor Cores
280
224
RT Cores
70
—
L1 Cache
128 KB (per SM)
192 KB (per SM)
L2 Cache
48 MB
24 MB
Theoretical Performance
Pixel Rate
137.5 GPixel/s
138.2 GPixel/s
Texture Rate
401.0 GTexel/s
322.6 GTexel/s
FP16 (half)
25.66 TFLOPS
10.32 TFLOPS
FP32 (float)
25.66 TFLOPS
10.32 TFLOPS
FP64 (double)
401.0 GFLOPS
5.161 TFLOPS
Graphics Processor
GPU Name
GB203
GA100
Architecture
Blackwell 2.0
Ampere
Foundry
TSMC
TSMC
Process Size
5 nm
7 nm
Transistors
45.6 billion
54.2 billion
Die Size
378 mm²
826 mm²
Board Design
TDP
70W
165W
Suggested PSU
250 W
450 W
Outputs
4x mini-DisplayPort 2.1b
No outputs
Power Connectors
None
8-pin EPS
Graphics Features
DirectX
12 Ultimate (12_2)
N/A
OpenGL
4.6
N/A
OpenCL
3.0
3.0
Vulkan
1.4
N/A
CUDA
12.0
8.0
Shader Model
6.8
N/A
▲ Specs Compare
Data sourced from official specifications and benchmark tests.