NVIDIA PG506 232 vs NVIDIA B100
We compared two Desktop platform GPUs: 24GB VRAM PG506 232 and 96GB VRAM B100 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA PG506 232 's Advantages
Lower TDP (165W vs 1000W)
NVIDIA B100 's Advantages
Released 3 years and 7 months late
Boost Clock has increased by 28% (1837MHz vs 1440MHz)
More VRAM (96GB vs 24GB)
Larger VRAM bandwidth (4.10TB/s vs 933.1GB/s)
13312 additional rendering cores
Score
Benchmark
FP32 (float)
PG506 232
10.32 TFLOPS
B100
+501%
62.08 TFLOPS
Graphics Card
Apr 2021
Release Date
Nov 2024
Tesla
Generation
Server Blackwell
Desktop
Type
Desktop
PCIe 4.0 x16
Bus Interface
PCIe 5.0 x16
Clock Speeds
930 MHz
Base Clock
1665 MHz
1440 MHz
Boost Clock
1837 MHz
1215 MHz
Memory Clock
2000 MHz
Memory
24GB
Memory Size
96GB
HBM2
Memory Type
HBM3e
3072bit
Memory Bus
4096bit
933.1GB/s
Bandwidth
4.10TB/s
Render Config
-
-
-
56
SM Count
132
3584
Shading Units
16896
224
TMUs
528
96
ROPs
24
224
Tensor Cores
528
-
-
-
192 KB (per SM)
L1 Cache
256 KB (per SM)
24 MB
L2 Cache
50 MB
-
-
-
Theoretical Performance
138.2 GPixel/s
Pixel Rate
44.09 GPixel/s
322.6 GTexel/s
Texture Rate
969.9 GTexel/s
10.32 TFLOPS
FP16 (half)
248.3 TFLOPS
10.32 TFLOPS
FP32 (float)
62.08 TFLOPS
5.161 TFLOPS
FP64 (double)
31.04 TFLOPS
Board Design
165W
TDP
1000W
450 W
Suggested PSU
1400 W
No outputs
Outputs
No outputs
8-pin EPS
Power Connectors
-
Graphics Processor
GA100
GPU Name
GB102
-
-
-
Ampere
Architecture
Blackwell
TSMC
Foundry
TSMC
7 nm
Process Size
5 nm
54.2 billion
Transistors
104 billion
826 mm²
Die Size
Unknown
Graphics Features
N/A
DirectX
N/A
N/A
OpenGL
N/A
3.0
OpenCL
3.0
N/A
Vulkan
N/A
8.0
CUDA
10.1
N/A
Shader Model
N/A
