NVIDIA L40 CNX vs NVIDIA Quadro K5200
We compared two Professional market GPUs: 24GB VRAM L40 CNX and 8GB VRAM Quadro K5200 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA L40 CNX 's Advantages
Released 8 years and 3 months late
Boost Clock has increased by 221% (2475MHz vs 771MHz)
More VRAM (24GB vs 8GB)
Larger VRAM bandwidth (864.0GB/s vs 192.3GB/s)
15872 additional rendering cores
NVIDIA Quadro K5200 's Advantages
Lower TDP (150W vs 300W)
Score
Benchmark
FP32 (float)
L40 CNX
+2432%
89.97 TFLOPS
Quadro K5200
3.553 TFLOPS
Graphics Card
Oct 2022
Release Date
Jul 2014
Tesla Ada
Generation
Quadro
Professional
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 3.0 x16
Clock Speeds
1005 MHz
Base Clock
667 MHz
2475 MHz
Boost Clock
771 MHz
2250 MHz
Memory Clock
1502 MHz
Memory
24GB
Memory Size
8GB
GDDR6
Memory Type
GDDR5
384bit
Memory Bus
256bit
864.0GB/s
Bandwidth
192.3GB/s
Render Config
-
-
-
142
SM Count
-
18176
Shading Units
2304
568
TMUs
192
192
ROPs
48
568
Tensor Cores
-
142
RT Cores
-
128 KB (per SM)
L1 Cache
-
48 MB
L2 Cache
-
-
-
-
Theoretical Performance
475.2 GPixel/s
Pixel Rate
37.01 GPixel/s
1406 GTexel/s
Texture Rate
148.0 GTexel/s
89.97 TFLOPS
FP16 (half)
-
89.97 TFLOPS
FP32 (float)
3.553 TFLOPS
1406 GFLOPS
FP64 (double)
148.0 GFLOPS
Board Design
300W
TDP
150W
700 W
Suggested PSU
450 W
1x HDMI 2.1
3x DisplayPort 1.4a
Outputs
2x DVI
2x DisplayPort 1.2
1x 16-pin
Power Connectors
1x 6-pin
Graphics Processor
AD102
GPU Name
GK110B
AD102-???-A1
GPU Variant
-
Ada Lovelace
Architecture
Kepler
TSMC
Foundry
TSMC
4 nm
Process Size
28 nm
76.3 billion
Transistors
7.08 billion
608 mm²
Die Size
561 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (11_1)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.1
8.9
CUDA
3.5
6.6
Shader Model
5.1
