NVIDIA B100 vs NVIDIA CMP 170HX
We compared two Desktop platform GPUs: 96GB VRAM B100 and 16GB VRAM CMP 170HX 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 2 months late
Boost Clock has increased by 30% (1837MHz vs 1410MHz)
More VRAM (96GB vs 16GB)
Larger VRAM bandwidth (4.10TB/s vs 1493GB/s)
12416 additional rendering cores
NVIDIA CMP 170HX 's Advantages
Lower TDP (250W vs 1000W)
Score
Benchmark
FP32 (float)
B100
+391%
62.08 TFLOPS
CMP 170HX
12.63 TFLOPS
Graphics Card
Nov 2024
Release Date
Sep 2021
Server Blackwell
Generation
Mining GPUs
Desktop
Type
Desktop
PCIe 5.0 x16
Bus Interface
PCIe 4.0 x4
Clock Speeds
1665 MHz
Base Clock
1140 MHz
1837 MHz
Boost Clock
1410 MHz
2000 MHz
Memory Clock
1458 MHz
Memory
96GB
Memory Size
16GB
HBM3e
Memory Type
HBM2e
4096bit
Memory Bus
4096bit
4.10TB/s
Bandwidth
1493GB/s
Render Config
-
-
-
132
SM Count
70
16896
Shading Units
4480
528
TMUs
280
24
ROPs
128
528
Tensor Cores
280
-
-
-
256 KB (per SM)
L1 Cache
192 KB (per SM)
50 MB
L2 Cache
8 MB
-
-
-
Theoretical Performance
44.09 GPixel/s
Pixel Rate
180.5 GPixel/s
969.9 GTexel/s
Texture Rate
394.8 GTexel/s
248.3 TFLOPS
FP16 (half)
50.53 TFLOPS
62.08 TFLOPS
FP32 (float)
12.63 TFLOPS
31.04 TFLOPS
FP64 (double)
6.317 TFLOPS
Board Design
1000W
TDP
250W
1400 W
Suggested PSU
600 W
No outputs
Outputs
No outputs
-
Power Connectors
2x 8-pin
Graphics Processor
GB102
GPU Name
GA100
-
GPU Variant
GA100-105F-A1
Blackwell
Architecture
Ampere
TSMC
Foundry
TSMC
5 nm
Process Size
7 nm
104 billion
Transistors
54.2 billion
Unknown
Die Size
826 mm²
Graphics Features
N/A
DirectX
N/A
N/A
OpenGL
N/A
3.0
OpenCL
3.0
N/A
Vulkan
N/A
10.1
CUDA
8.0
N/A
Shader Model
N/A
