NVIDIA GRID M3 3020 vs NVIDIA RTX 4000 SFF Ada Generation
We compared two Professional market GPUs: 4GB VRAM GRID M3 3020 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 RTX 4000 SFF Ada Generation 's Advantages
Released 2 years and 6 months late
Boost Clock has increased by 34% (1746MHz vs 1306MHz)
More VRAM (16GB vs 4GB)
Larger VRAM bandwidth (1.02TB/s vs 83.20GB/s)
3200 additional rendering cores
Score
Benchmark
FP32 (float)
GRID M3 3020
1.672 TFLOPS
RTX 4000 SFF Ada Generation
+702%
13.41 TFLOPS
Graphics Card
May 2016
Release Date
Nov 2018
GRID
Generation
Radeon Instinct
Professional
Type
Professional
PCIe 3.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
1033 MHz
Base Clock
1200 MHz
1306 MHz
Boost Clock
1746 MHz
1300 MHz
Memory Clock
1000 MHz
Memory
4GB
Memory Size
16GB
GDDR5
Memory Type
HBM2
128bit
Memory Bus
4096bit
83.20GB/s
Bandwidth
1.02TB/s
Render Config
-
Compute Units
60
-
SM Count
48
640
Shading Units
3840
40
TMUs
240
16
ROPs
64
-
Tensor Cores
192
-
RT Cores
48
64 KB (per SMM)
L1 Cache
16 KB (per CU)
2 MB
L2 Cache
4 MB
-
-
-
Theoretical Performance
20.90 GPixel/s
Pixel Rate
111.7 GPixel/s
52.24 GTexel/s
Texture Rate
419.0 GTexel/s
-
FP16 (half)
26.82 TFLOPS
1.672 TFLOPS
FP32 (float)
13.41 TFLOPS
52.24 GFLOPS
FP64 (double)
6.705 TFLOPS
Board Design
Unknown
TDP
300W
200 W
Suggested PSU
700 W
No outputs
Outputs
1x mini-DisplayPort 1.4a
1x 8-pin
Power Connectors
2x 8-pin
Graphics Processor
GM107
GPU Name
Vega 20
-
GPU Variant
Vega 20 GLXT
Maxwell
Architecture
GCN 5.1
TSMC
Foundry
TSMC
28 nm
Process Size
7 nm
1.87 billion
Transistors
13.23 billion
148 mm²
Die Size
331 mm²
Graphics Features
12 (11_0)
DirectX
12 (12_1)
4.6
OpenGL
4.6
3.0
OpenCL
2.1
1.3
Vulkan
1.3
5.0
CUDA
8.9
6.7 (5.1)
Shader Model
6.7
