NVIDIA A40 PCIe vs NVIDIA L20
GPU · Specs Compare
Graphics Card
Release Date
Oct 2020
Nov 2023
Generation
Tesla
Server Ada
Type
Desktop
Desktop
Bus Interface
PCIe 4.0 x16
PCIe 4.0 x16
Clock Speeds
Base Clock
1305 MHz
1440 MHz
Boost Clock
1740 MHz
2520 MHz
Memory Clock
1812 MHz
2250 MHz
Memory
Memory Size
48GB
48GB
Memory Type
GDDR6
GDDR6
Memory Bus
384bit
384bit
Bandwidth
695.8GB/s
864.0GB/s
Render Config
SM Count
84
92
Shading Units
10752
11776
TMUs
336
368
ROPs
112
128
Tensor Cores
336
368
RT Cores
84
92
L1 Cache
128 KB (per SM)
128 KB (per SM)
L2 Cache
6 MB
96 MB
Theoretical Performance
Pixel Rate
194.9 GPixel/s
322.6 GPixel/s
Texture Rate
584.6 GTexel/s
927.4 GTexel/s
FP16 (half)
37.42 TFLOPS
59.35 TFLOPS
FP32 (float)
37.42 TFLOPS
59.35 TFLOPS
FP64 (double)
584.6 GFLOPS
927.4 GFLOPS
Graphics Processor
GPU Name
GA102
AD102
Architecture
Ampere
Ada Lovelace
Foundry
Samsung
TSMC
Process Size
8 nm
5 nm
Transistors
28.3 billion
76.3 billion
Die Size
628 mm²
609 mm²
GPU Variant
—
AD102
Board Design
TDP
300W
275W
Suggested PSU
700 W
600 W
Outputs
3x DisplayPort 1.4a
4x DisplayPort 1.4a
Power Connectors
8-pin EPS
1x 16-pin
Graphics Features
DirectX
12 Ultimate (12_2)
12 Ultimate (12_2)
OpenGL
4.6
4.6
OpenCL
3.0
3.0
Vulkan
1.3
1.4
CUDA
8.6
8.9
Shader Model
6.6
6.9
▲ Specs Compare
Data sourced from official specifications and benchmark tests.