NVIDIA Jetson Orin NX 16GB (MaxQ): edge AI benchmarks

Vendor NVIDIAForm factor edge moduleCompute GPU+DLAIndicative price unknownPower envelope unknownPublished rows 5

Key results

Best published result per workload for NVIDIA Jetson Orin NX 16GB (MaxQ)
Workload	Best result	Model · precision	Power · perf/W	Grade · date	Source
object detection	47.59 FPS batch unknown	MLPerf Retinanet INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	15.57 W (measured) 3.06 FPS/W	D Older data (2023)	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table)
image classification	1682 FPS batch unknown	MLPerf ResNet INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	14.95 W (measured) 112 FPS/W	D Older data (2023)	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table)
speech-to-text	328 FPS batch unknown	MLPerf RNN-T INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	17.25 W (measured) 19.00 FPS/W	D Older data (2023)	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table)
language models (NLP)	137 FPS batch unknown	MLPerf BERT INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	17.04 W (measured) 8.02 FPS/W	D Older data (2023)	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table)
medical segmentation	0.190 FPS batch unknown	MLPerf 3D-Unet-99.0 INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	22.04 W (measured) 0.00862 FPS/W	D Older data (2023)	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table)

Best published result per workload, across model sizes. For like-for-like groups, open the per-workload tables below.

Show data — per-workload tables

object detection

Like-for-like group	Best result	Model · precision · stack	Grade	Power	Source · date	Rank in group
mid-size CNN · b=unknown	47.59 FPS	MLPerf Retinanet INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	D	15.57 W measured	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table) Older data (2023)	15 of 34

instance segmentation

◌

No published instance segmentation results for NVIDIA Jetson Orin NX 16GB (MaxQ) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

pose estimation

◌

No published pose estimation results for NVIDIA Jetson Orin NX 16GB (MaxQ) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

image classification

Like-for-like group	Best result	Model · precision · stack	Grade	Power	Source · date	Rank in group
mid-size CNN · b=unknown	1682 FPS	MLPerf ResNet INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	D	14.95 W measured	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table) Older data (2023)	20 of 41

semantic segmentation

◌

No published semantic segmentation results for NVIDIA Jetson Orin NX 16GB (MaxQ) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

speech-to-text

Like-for-like group	Best result	Model · precision · stack	Grade	Power	Source · date	Rank in group
RNN-T · b=unknown	328 FPS	MLPerf RNN-T INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	D	17.25 W measured	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table) Older data (2023)	3 of 3

LLM generation

◌

No published LLM generation results for NVIDIA Jetson Orin NX 16GB (MaxQ) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

language models (NLP)

Like-for-like group	Best result	Model · precision · stack	Grade	Power	Source · date	Rank in group
unknown · b=unknown	137 FPS	MLPerf BERT INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	D	17.04 W measured	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table) Older data (2023)	17 of 25

medical segmentation

Like-for-like group	Best result	Model · precision · stack	Grade	Power	Source · date	Rank in group
mid-size CNN · b=unknown	0.190 FPS	MLPerf 3D-Unet-99.0 INT8/FP16 (MLPerf closed) · TensorRT (MLPerf Inference, older round)	D	22.04 W measured	NVIDIA Jetson Benchmarks (MLPerf v3.1 results table) Older data (2023)	4 of 4

How to read these numbers

Every number is labelled. Confidence: A MLPerf closed, audited · B documented third-party method · C community / review · D vendor-published, not independently verified · E estimate. Power basis: measured with the result · spec vendor spec/envelope · power n/a not published.

Machine-readable: /find.json (see llms.txt); full tables are in “Show data”. Confidence grades: A audited MLPerf closed · B documented third-party · C community · D vendor-published · E estimate. Unknown values are shown as unknown, never estimated. Generated 2026-10-16 from public sources.