NVIDIA Quadro K620 vs NVIDIA RTX A500 Embedded
We compared two Professional market GPUs: 2GB VRAM Quadro K620 and 4GB VRAM RTX A500 Embedded to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA RTX A500 Embedded 's Advantages
Released 7 years and 8 months late
Boost Clock has increased by 19% (1335MHz vs 1124MHz)
More VRAM (4GB vs 2GB)
Larger VRAM bandwidth (96.00GB/s vs 28.80GB/s)
1664 additional rendering cores
Lower TDP (20W vs 45W)
Score
Benchmark
FP32 (float)
Quadro K620
0.863 TFLOPS
RTX A500 Embedded
+533%
5.468 TFLOPS
3DMark Time Spy
Quadro K620
696
RTX A500 Embedded
+277%
2628
Graphics Card
Jul 2014
Release Date
Mar 2022
Quadro
Generation
Quadro Ampere-M
Professional
Type
Professional
PCIe 2.0 x16
Bus Interface
PCIe 4.0 x8
Clock Speeds
1058 MHz
Base Clock
435 MHz
1124 MHz
Boost Clock
1335 MHz
900 MHz
Memory Clock
1500 MHz
Memory
2GB
Memory Size
4GB
DDR3
Memory Type
GDDR6
128bit
Memory Bus
64bit
28.80GB/s
Bandwidth
96.00GB/s
Render Config
-
-
-
-
SM Count
16
384
Shading Units
2048
24
TMUs
64
16
ROPs
32
-
Tensor Cores
64
-
RT Cores
16
64 KB (per SMM)
L1 Cache
128 KB (per SM)
2 MB
L2 Cache
2 MB
-
-
-
Theoretical Performance
17.98 GPixel/s
Pixel Rate
42.72 GPixel/s
26.98 GTexel/s
Texture Rate
85.44 GTexel/s
-
FP16 (half)
5.468 TFLOPS
863.2 GFLOPS
FP32 (float)
5.468 TFLOPS
26.98 GFLOPS
FP64 (double)
85.44 GFLOPS
Board Design
45W
TDP
20W
200 W
Suggested PSU
-
1x DVI
1x DisplayPort 1.4a
Outputs
Portable Device Dependent
None
Power Connectors
None
Graphics Processor
GM107
GPU Name
GA107S
GM107-850-A2
GPU Variant
-
Maxwell
Architecture
Ampere
TSMC
Foundry
Samsung
28 nm
Process Size
8 nm
1.87 billion
Transistors
8.7 billion
148 mm²
Die Size
200 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
