NVIDIA B100 vs NVIDIA CMP 40HX
We compared two Desktop platform GPUs: 96GB VRAM B100 and 8GB VRAM CMP 40HX to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA B100 's Advantages
Released 3 years and 9 months late
Boost Clock has increased by 11% (1837MHz vs 1650MHz)
More VRAM (96GB vs 8GB)
Larger VRAM bandwidth (4.10TB/s vs 448.0GB/s)
14592 additional rendering cores
NVIDIA CMP 40HX 's Advantages
Lower TDP (185W vs 1000W)
Score
Benchmark
FP32 (float)
B100
+716%
62.08 TFLOPS
CMP 40HX
7.603 TFLOPS
Graphics Card
Nov 2024
Release Date
Feb 2021
Server Blackwell
Generation
Mining GPUs
Desktop
Type
Desktop
PCIe 5.0 x16
Bus Interface
PCIe 1.0 x4
Clock Speeds
1665 MHz
Base Clock
1470 MHz
1837 MHz
Boost Clock
1650 MHz
2000 MHz
Memory Clock
1750 MHz
Memory
96GB
Memory Size
8GB
HBM3e
Memory Type
GDDR6
4096bit
Memory Bus
256bit
4.10TB/s
Bandwidth
448.0GB/s
Render Config
-
-
-
132
SM Count
36
16896
Shading Units
2304
528
TMUs
144
24
ROPs
64
528
Tensor Cores
288
-
RT Cores
36
256 KB (per SM)
L1 Cache
64 KB (per SM)
50 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
44.09 GPixel/s
Pixel Rate
105.6 GPixel/s
969.9 GTexel/s
Texture Rate
237.6 GTexel/s
248.3 TFLOPS
FP16 (half)
15.21 TFLOPS
62.08 TFLOPS
FP32 (float)
7.603 TFLOPS
31.04 TFLOPS
FP64 (double)
237.6 GFLOPS
Board Design
1000W
TDP
185W
1400 W
Suggested PSU
450 W
No outputs
Outputs
No outputs
-
Power Connectors
1x 8-pin
Graphics Processor
GB102
GPU Name
TU106
-
GPU Variant
TU106-100-A1
Blackwell
Architecture
Turing
TSMC
Foundry
TSMC
5 nm
Process Size
12 nm
104 billion
Transistors
10.8 billion
Unknown
Die Size
445 mm²
Graphics Features
N/A
DirectX
12 Ultimate (12_2)
N/A
OpenGL
4.6
3.0
OpenCL
3.0
N/A
Vulkan
1.3
10.1
CUDA
7.5
N/A
Shader Model
6.8
