GRID M40 vs Data Center GPU Max Subsystem
We compared two Professional market GPUs: 8GB VRAM GRID M40 and 128GB VRAM Data Center GPU Max Subsystem 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 2400W)
Data Center GPU Max Subsystem 's Advantages
Released 6 years and 8 months late
Boost Clock1600MHz
More VRAM (128GB vs 8GB)
Larger VRAM bandwidth (3205GB/s vs 83.20GB/s)
16000 additional rendering cores
Score
Benchmark
FP32 (float)
GRID M40
793
Data Center GPU Max Subsystem
+6511%
52430
Graphics Card
May 2016
Release Date
Jan 2023
GRID
Generation
Data Center GPU
Professional
Type
Professional
PCIe 3.0 x16
Bus Interface
PCIe 5.0 x16
Clock Speeds
-
Base Clock
900 MHz
-
Boost Clock
1600 MHz
1300 MHz
Memory Clock
1565 MHz
Memory
8GB
Memory Size
128GB
GDDR5
Memory Type
HBM2e
128bit
Memory Bus
8192bit
83.20GB/s
Bandwidth
3205GB/s
Render Config
384
Shading Units
16384
32
TMUs
1024
16
ROPs
0
-
Tensor Cores
1024
-
RT Cores
128
64 KB (per SMM)
L1 Cache
64 KB (per EU)
2 MB
L2 Cache
408 MB
Theoretical Performance
16.53 GPixel/s
Pixel Rate
0 MPixel/s
33.06 GTexel/s
Texture Rate
1638 GTexel/s
-
FP16 (half)
52.43 TFLOPS
793.3 GFLOPS
FP32 (float)
52.43 TFLOPS
24.79 GFLOPS
FP64 (double)
52.43 TFLOPS
Board Design
50W
TDP
2400W
250 W
Suggested PSU
2800 W
No outputs
Outputs
No outputs
-
Power Connectors
1x 16-pin
Graphics Processor
GM107
GPU Name
Ponte Vecchio
Maxwell
Architecture
Generation 12.5
TSMC
Foundry
Intel
28 nm
Process Size
10 nm
1.87 billion
Transistors
100 billion
148 mm²
Die Size
1280 mm²
Graphics Features
12 (11_0)
DirectX
12 (12_1)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
1.3
Vulkan
N/A
5.0
CUDA
-
5.1
Shader Model
6.6
