Data Center GPU Flex 170 vs GRID A100B
We compared two Professional market GPUs: 16GB VRAM Data Center GPU Flex 170 and 48GB VRAM GRID A100B to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
Data Center GPU Flex 170 's Advantages
Released 2 years and 3 months late
Boost Clock has increased by 104% (2050MHz vs 1005MHz)
Lower TDP (150W vs 400W)
GRID A100B 's Advantages
More VRAM (48GB vs 16GB)
Larger VRAM bandwidth (1866GB/s vs 512.0GB/s)
2816 additional rendering cores
Score
Benchmark
FP32 (float)
Data Center GPU Flex 170
+20%
16790
GRID A100B
13890
Graphics Card
Aug 2022
Release Date
May 2020
Data Center GPU
Generation
GRID
Professional
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
1950 MHz
Base Clock
900 MHz
2050 MHz
Boost Clock
1005 MHz
2000 MHz
Memory Clock
1215 MHz
Memory
16GB
Memory Size
48GB
GDDR6
Memory Type
HBM2e
256bit
Memory Bus
6144bit
512.0GB/s
Bandwidth
1866GB/s
Render Config
-
SM Count
108
4096
Shading Units
6912
256
TMUs
432
128
ROPs
192
-
Tensor Cores
432
32
RT Cores
-
-
L1 Cache
192 KB (per SM)
16 MB
L2 Cache
48 MB
Theoretical Performance
262.4 GPixel/s
Pixel Rate
193.0 GPixel/s
524.8 GTexel/s
Texture Rate
434.2 GTexel/s
33.59 TFLOPS
FP16 (half)
55.57 TFLOPS
16.79 TFLOPS
FP32 (float)
13.89 TFLOPS
-
FP64 (double)
6.947 TFLOPS
Board Design
150W
TDP
400W
450 W
Suggested PSU
800 W
1x HDMI 2.1
3x DisplayPort 2.0
Outputs
No outputs
1x 8-pin
Power Connectors
None
Graphics Processor
DG2-512
GPU Name
GA100
ACM-G10
GPU Variant
-
Generation 12.7
Architecture
Ampere
TSMC
Foundry
TSMC
6 nm
Process Size
7 nm
21.7 billion
Transistors
54.2 billion
406 mm²
Die Size
826 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
N/A
4.6
OpenGL
N/A
3.0
OpenCL
3.0
1.3
Vulkan
N/A
-
CUDA
8.0
6.6
Shader Model
N/A
