NVIDIA RTX PRO 6000 Blackwell Server vs NVIDIA H100 PCIe 80 GB

GPU Specs Compare

Graphics Card

Release Date
Mar 2025
Oct 2022
Generation
Server Blackwell
Server Hopper
Type
Professional
AI GPU
Bus Interface
PCIe 5.0 x16
PCIe 5.0 x16

Clock Speeds

Base Clock
1590 MHz
1095 MHz
Boost Clock
2617 MHz
1755 MHz
Memory Clock
1750 MHz
1593 MHz

Memory

Memory Size
96GB
80GB
Memory Type
GDDR7
HBM2e
Memory Bus
512bit
5120bit
Bandwidth
1.79TB/s
2.04TB/s

Render Config

SM Count
188
114
Shading Units
24064
14592
TMUs
752
456
ROPs
192
24
Tensor Cores
752
456
RT Cores
188
-
L1 Cache
128 KB (per SM)
256 KB (per SM)
L2 Cache
128 MB
50 MB

Theoretical Performance

Pixel Rate
502.5 GPixel/s
42.12 GPixel/s
Texture Rate
1968.0 GTexel/s
800.3 GTexel/s
FP16 (half)
126.0 TFLOPS
204.9 TFLOPS
FP32 (float)
126.0 TFLOPS
51.22 TFLOPS
FP64 (double)
1.968 TFLOPS
25.61 TFLOPS

Graphics Processor

GPU Name
GB202
GH100
Architecture
Blackwell 2.0
Hopper
Foundry
TSMC
TSMC
Process Size
5 nm
5 nm
Transistors
92.2 billion
80 billion
Die Size
750 mm²
814 mm²

Board Design

TDP
600W
350W
Suggested PSU
1000 W
750 W
Outputs
4x DisplayPort 2.1b
No outputs
Power Connectors
1x 16-pin
1x 16-pin

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
9.0
Shader Model
6.9
N/A