NVIDIA L40 vs NVIDIA RTX 4000 SFF Ada Generation
We compared two Professional market GPUs: 48GB VRAM L40 and 16GB VRAM RTX 4000 SFF Ada Generation to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA L40 's Advantages
Released 3 years and 11 months late
Boost Clock has increased by 43% (2490MHz vs 1746MHz)
More VRAM (48GB vs 16GB)
14336 additional rendering cores
NVIDIA RTX 4000 SFF Ada Generation 's Advantages
Larger VRAM bandwidth (1.02TB/s vs 864.0GB/s)
Score
Benchmark
FP32 (float)
L40
+575%
90.52 TFLOPS
RTX 4000 SFF Ada Generation
13.41 TFLOPS
Blender
L40
+4%
3901
RTX 4000 SFF Ada Generation
3717
Graphics Card
Oct 2022
Release Date
Nov 2018
Tesla Ada
Generation
Radeon Instinct
Professional
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
735 MHz
Base Clock
1200 MHz
2490 MHz
Boost Clock
1746 MHz
2250 MHz
Memory Clock
1000 MHz
Memory
48GB
Memory Size
16GB
GDDR6
Memory Type
HBM2
384bit
Memory Bus
4096bit
864.0GB/s
Bandwidth
1.02TB/s
Render Config
-
Compute Units
60
142
SM Count
48
18176
Shading Units
3840
568
TMUs
240
192
ROPs
64
568
Tensor Cores
192
142
RT Cores
48
128 KB (per SM)
L1 Cache
16 KB (per CU)
96 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
478.1 GPixel/s
Pixel Rate
111.7 GPixel/s
1414 GTexel/s
Texture Rate
419.0 GTexel/s
90.52 TFLOPS
FP16 (half)
26.82 TFLOPS
90.52 TFLOPS
FP32 (float)
13.41 TFLOPS
1414 GFLOPS
FP64 (double)
6.705 TFLOPS
Board Design
300W
TDP
300W
700 W
Suggested PSU
700 W
4x DisplayPort 1.4a
Outputs
1x mini-DisplayPort 1.4a
1x 16-pin
Power Connectors
2x 8-pin
Graphics Processor
AD102
GPU Name
Vega 20
AD102-895-A1
GPU Variant
Vega 20 GLXT
Ada Lovelace
Architecture
GCN 5.1
TSMC
Foundry
TSMC
4 nm
Process Size
7 nm
76.3 billion
Transistors
13.23 billion
608 mm²
Die Size
331 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (12_1)
4.6
OpenGL
4.6
3.0
OpenCL
2.1
1.3
Vulkan
1.3
8.9
CUDA
8.9
6.6
Shader Model
6.7
