NVIDIA RTX PRO 5000 Blackwell vs AMD Radeon E8870

GPU · Specs Compare

Products
GPU NVIDIA RTX PRO 5000 Blackwell NVIDIA · 2025-03
GPU AMD Radeon E8870 AMD · 2015-09
Graphics Card
Release Date
Mar 2025
Sep 2015
Generation
Blackwell PRO W
Embedded
Type
Desktop
Mobile
Bus Interface
PCIe 5.0 x16
MXM-B (3.0)
Clock Speeds
Base Clock
1740 MHz
—
Boost Clock
2377 MHz
—
Memory Clock
1750 MHz
1500 MHz
Memory
Memory Size
48GB
4GB
Memory Type
GDDR7
GDDR5
Memory Bus
384bit
128bit
Bandwidth
1.34TB/s
96.00GB/s
Render Config
SM Count
110
—
Shading Units
14080
768
TMUs
440
48
ROPs
160
16
Tensor Cores
440
—
RT Cores
110
—
L1 Cache
128 KB (per SM)
16 KB (per CU)
L2 Cache
96 MB
256 KB
Compute Units
—
12
Theoretical Performance
Pixel Rate
380.3 GPixel/s
16.00 GPixel/s
Texture Rate
1045.9 GTexel/s
48.00 GTexel/s
FP16 (half)
66.94 TFLOPS
—
FP32 (float)
66.94 TFLOPS
1.536 TFLOPS
FP64 (double)
1045.9 GFLOPS
96.00 GFLOPS
Graphics Processor
GPU Name
GB202
Saturn
Architecture
Blackwell 2.0
GCN 2.0
Foundry
TSMC
TSMC
Process Size
5 nm
28 nm
Transistors
92.2 billion
2.08 billion
Die Size
750 mm²
160 mm²
GPU Variant
—
Saturn PRO
Board Design
TDP
300W
75W
Suggested PSU
700 W
—
Outputs
4x DisplayPort 2.1b
No outputs
Power Connectors
1x 16-pin
None
Graphics Features
DirectX
12 Ultimate (12_2)
12 (12_0)
OpenGL
4.6
4.6
OpenCL
3.0
2.0
Vulkan
1.4
1.2
CUDA
12.0
—
Shader Model
6.9
6.3
Benchmark
3DMark Time Spy
33,196
—
3DMark Time Spy Extreme
17,934
—
3DMark Speed Way
9,504
—
▲ Specs Compare Data sourced from official specifications and benchmark tests.