NVIDIA L4 vs NVIDIA GRID M40
We compared two Professional market GPUs: 24GB VRAM L4 and 8GB VRAM GRID M40 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA L4 's Advantages
Released 6 years and 10 months late
Boost Clock2040MHz
More VRAM (24GB vs 8GB)
Larger VRAM bandwidth (300.1GB/s vs 83.20GB/s)
7040 additional rendering cores
NVIDIA GRID M40 's Advantages
Lower TDP (50W vs 72W)
Score
Benchmark
FP32 (float)
L4
+3719%
30.29 TFLOPS
GRID M40
0.793 TFLOPS
Graphics Card
Mar 2023
Release Date
May 2016
Tesla Ada(Lxx)
Generation
GRID
Professional
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 3.0 x16
Clock Speeds
795 MHz
Base Clock
-
2040 MHz
Boost Clock
-
1563 MHz
Memory Clock
1300 MHz
Memory
24GB
Memory Size
8GB
GDDR6
Memory Type
GDDR5
192bit
Memory Bus
128bit
300.1GB/s
Bandwidth
83.20GB/s
Render Config
-
-
-
60
SM Count
-
7424
Shading Units
384
240
TMUs
32
80
ROPs
16
240
Tensor Cores
-
60
RT Cores
-
128 KB (per SM)
L1 Cache
64 KB (per SMM)
48 MB
L2 Cache
2 MB
-
-
-
Theoretical Performance
163.2 GPixel/s
Pixel Rate
16.53 GPixel/s
489.6 GTexel/s
Texture Rate
33.06 GTexel/s
30.29 TFLOPS
FP16 (half)
-
30.29 TFLOPS
FP32 (float)
793.3 GFLOPS
473.3 GFLOPS
FP64 (double)
24.79 GFLOPS
Board Design
72W
TDP
50W
250 W
Suggested PSU
250 W
No outputs
Outputs
No outputs
None
Power Connectors
-
Graphics Processor
AD104
GPU Name
GM107
AD104-???-A1
GPU Variant
-
Ada Lovelace
Architecture
Maxwell
TSMC
Foundry
TSMC
5 nm
Process Size
28 nm
35.8 billion
Transistors
1.87 billion
294 mm²
Die Size
148 mm²
Graphics Features
12 Ultimate (12_2)
DirectX
12 (11_0)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.3
8.9
CUDA
5.0
6.8
Shader Model
5.1
