NVIDIA RTX PRO 4000 Blackwell SFF vs ATI Radeon HD 5870
GPU · Specs Compare
Graphics Card
Release Date
Aug 2025
Sep 2009
Generation
Blackwell PRO W
Evergreen
Type
Desktop
Desktop
Bus Interface
PCIe 5.0 x8
PCIe 2.0 x16
Clock Speeds
Base Clock
577 MHz
—
Boost Clock
1432 MHz
—
Memory Clock
1125 MHz
1200 MHz
Memory
Memory Size
24GB
1024MB
Memory Type
GDDR7
GDDR5
Memory Bus
192bit
256bit
Bandwidth
432.0GB/s
153.6GB/s
Render Config
SM Count
70
—
Shading Units
8960
1600
TMUs
280
80
ROPs
96
32
Tensor Cores
280
—
RT Cores
70
—
L1 Cache
128 KB (per SM)
8 KB (per CU)
L2 Cache
48 MB
512 KB
Compute Units
—
20
Theoretical Performance
Pixel Rate
137.5 GPixel/s
27.20 GPixel/s
Texture Rate
401.0 GTexel/s
68.00 GTexel/s
FP16 (half)
25.66 TFLOPS
—
FP32 (float)
25.66 TFLOPS
2.720 TFLOPS
FP64 (double)
401.0 GFLOPS
544.0 GFLOPS
Graphics Processor
GPU Name
GB203
Cypress
Architecture
Blackwell 2.0
TeraScale 2
Foundry
TSMC
TSMC
Process Size
5 nm
40 nm
Transistors
45.6 billion
2.154 billion
Die Size
378 mm²
334 mm²
GPU Variant
—
Cypress XT(215-0735033)
Board Design
TDP
70W
188W
Suggested PSU
250 W
450 W
Outputs
4x mini-DisplayPort 2.1b
2x DVI1x HDMI 1.3a1x DisplayPort 1.1
Power Connectors
None
2x 6-pin
Graphics Features
DirectX
12 Ultimate (12_2)
11.2 (11_0)
OpenGL
4.6
4.4
OpenCL
3.0
1.2
Vulkan
1.4
N/A
CUDA
12.0
—
Shader Model
6.8
5.0
▲ Specs Compare
Data sourced from official specifications and benchmark tests.