89
edits
Changes
no edit summary
!Model
!Threads parameter
!Prediction time
[ms]
| rowspan="3" |Floating-point
|unspecified
|220
|-
|1
|220
|-
|2
|390
|-
|Half-quantized
|unspecified
|330
|-
| rowspan="2" |Fully-quantized
|unspecified
|200
|-
|4
|84
|-
| rowspan="5" |'''Xilinx Zynq UltraScale+ MPSoC ZCU104 Evaluation Kit'''
|2 cores DPU (through DNNDK API)
|Fully-quantized US+ (*)
|
|3.7
|-
| rowspan="4" |2 cores DPU (through VART API)
| rowspan="4" |Fully-quantized US+ (*)
|1
|4.1
|-
|2
|2.3
|-
|4
|1.2
|-
|6
|1.2
|-
| rowspan="7" |'''NXP i.MX8M Plus EVK'''
| rowspan="3" |Floating-point
|unspecified
|89
|-
|1
|160
|-
|2
|130
|-
|Half-quantized
|unspecified
|180
|-
| rowspan="2" |Fully-quantized
|unspecified
|85
|-
|4
|29
|-
|Fully-quantized
|NA
|1.5
|}
(*) This fully-quantized model differs from the one tested with the other two platforms because of the different [https://wiki.dave.eu/index.php/ML-TN-001_-_AI_at_the_edge:_comparison_of_different_embedded_platforms_-_Part_3#Building_the_application workflow needed to deploy it]. In particular, it is not in the TFL format. Anyway, both fully-quantized models were obtained from TensorFlow models sharing the same structure, as every other model listed.
