GRID M40 vs RTX A4000
We compared two Professional market GPUs: 8GB VRAM GRID M40 and 16GB VRAM RTX A4000 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
GRID M40 's Advantages
Lower TDP (50W vs 140W)
RTX A4000 's Advantages
Released 4 years and 11 months late
Boost Clock1560MHz
More VRAM (16GB vs 8GB)
Larger VRAM bandwidth (448.0GB/s vs 83.20GB/s)
5760 additional rendering cores
Score
Benchmark
FP32 (float)
GRID M40
793
RTX A4000
+2317%
19170
Graphics Card
May 2016
Release Date
Apr 2021
GRID
Generation
Quadro Ampere
Professional
Type
Professional
PCIe 3.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
-
Base Clock
735 MHz
-
Boost Clock
1560 MHz
1300 MHz
Memory Clock
1750 MHz
Memory
8GB
Memory Size
16GB
GDDR5
Memory Type
GDDR6
128bit
Memory Bus
256bit
83.20GB/s
Bandwidth
448.0GB/s
Render Config
-
SM Count
48
384
Shading Units
6144
32
TMUs
192
16
ROPs
96
-
Tensor Cores
192
-
RT Cores
48
64 KB (per SMM)
L1 Cache
128 KB (per SM)
2 MB
L2 Cache
4 MB
Theoretical Performance
16.53 GPixel/s
Pixel Rate
149.8 GPixel/s
33.06 GTexel/s
Texture Rate
299.5 GTexel/s
-
FP16 (half)
19.17 TFLOPS
793.3 GFLOPS
FP32 (float)
19.17 TFLOPS
24.79 GFLOPS
FP64 (double)
299.5 GFLOPS
Board Design
50W
TDP
140W
250 W
Suggested PSU
300 W
No outputs
Outputs
4x DisplayPort 1.4a
-
Power Connectors
1x 6-pin
Graphics Processor
GM107
GPU Name
GA104
-
GPU Variant
GA104-875-A1
Maxwell
Architecture
Ampere
TSMC
Foundry
Samsung
28 nm
Process Size
8 nm
1.87 billion
Transistors
17.4 billion
148 mm²
Die Size
392 mm²
Graphics Features
12 (11_0)
DirectX
12 Ultimate (12_2)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
1.3
5.0
CUDA
8.6
5.1
Shader Model
6.7
