Skip to content
ixian
Optimize this model
← Model library

GR00T N1.7

VLA · NVIDIA · measured on AGX Thor and AGX Orin

GR00T N1.7 robot foundation model: two camera frames and the robot state in, a decoded action chunk out.

GR00T N1.7 on Hugging Face ↗

Compare with other models on AGX Thor →

Performance highlights

PyTorch eager → ZETIC optimized · AGX Thor

Policy latency · p50
78.4%4.6× faster
Energy per inference
70.0%Less energy
Peak memory
34.9%Less memory

Performance comparison

Lower is better

AGX Thor · per policy call

Recorded measurements for the selected reference and ZETIC optimized paths.
MetricReferencePyTorch eager · BF16ZETIC optimizedMixed precision
Policy latency · p50108.52 ms23.46 ms
Energy per inference7.11 J2.13 J
Peak memory15.57 GB10.14 GB

View optimized build results →

Measurement conditions and source

Peak board memory above idle (MemTotal − MemAvailable), over the full command including model loading; not process RSS.

p50. GR00T N1.7 · LIBERO 10 · AGX Thor

Speedup is reference latency divided by path latency on this board. Lowest recorded latency excludes marked model or inference-schedule variants. Check the configurations below before treating paths as a matched comparison.

PyTorch eager · Model authors · BF16

PyTorch eager · BF16 · batch 1

Remeasurement pending. Peak board memory above idle (MemTotal − MemAvailable), over the full command including model loading; not process RSS.

ZETIC · ZETIC · Mixed precision

ZETIC · MIXED PRECISION · batch 1

Remeasurement pending. Quality validation pending. Additional validation pending. Peak board memory above idle (MemTotal − MemAvailable), over the full command including model loading; not process RSS.

PyTorch eager: End-to-end latency, p50 (median of round medians), 1 round · 553 samples per round.

ZETIC: End-to-end latency, p50 (median of round medians), 1 round · 2,553 samples per round.

Source: reviewed benchmark measurements. Public measurement: record-b5d142329cc468f9c335b93b6624bb9097f16eaf56323e517756d38cb8d404e3.

27/178 LIBERO-10 fixtures; one instruction (bucket 156).

Measurement methodology ↗

Quality evaluation

GR00T N1.7 · LIBERO 10

AGX Thor

PyTorch eager

BF16

95.02%191/201

Backend
PyTorch eager
Precision
BF16
Tasks
10
Episodes
201

ZETIC optimized

Precision not reported

94.55%191/202

Backend
ZETIC
Tasks
10
Episodes
202
Checkpoint
nvidia/GR00T-N1.7-LIBERO
Protocol
Not reported
Seeds
PyTorch eager · Seed not reported: 191/201ZETIC optimized · Seed not reported: 191/202
Source
Aggregate episode totals were reported; per-task outcomes were not supplied.

Evaluation coverage differs; no paired success-rate delta.

Evaluation details and episode results

Build with ZETIC

Optimize GR00T N1.7.

Share your GR00T N1.7 checkpoint for AGX Thor, with goals for lower latency, energy per inference and memory use.

We’ll reply by email.