Nvidia reports edge-agent benchmark result on Jetson AGX Thor

Image: Primary

Published Wednesday, September 16, 2026 · 8:07 PM EDT

Nvidia said its TensorRT Edge-LLM system completed the MLPerf Inference v6.1 Edge Agentic performance workload in 24 minutes and 36 seconds on one Jetson AGX Thor Developer Kit. The company reported 52.33 tokens per second running Qwen3.6-27B, compared with two hours and 37 minutes for the cited llama.cpp reference submission. Nvidia attributed the result to NVFP4 quantization, tree-based multi-token prediction and reuse of cached conversation state. The workload contained 20 software-engineering agent conversations and 1,007 generated turns.

Sources

NVIDIA Technical Blog

In this story

Nvidia

Published by Tech & Business, a media brand covering technology and business. This story was sourced from NVIDIA Technical Blog and reviewed by the T&B editorial agent team.

Back to Newswire

Keep reading

Full wire

AI Capital

Lumaril raises $66 million for industrial AI software

Lumaril raised a $66 million Series A at a $400 million valuation, The Wall Street Journal reported. The company develops AI software that connects to manufacturers' existing control systems to optimize industrial production on factory floors.

Capital

NeuroXess raises RMB 700 million for brain-computer interfaces

Shanghai-based NeuroXess raised RMB 700 million in Series B funding, TechNode reported. SDIC Juli, Shenzhen Capital Group and Qinzhi Capital co-led the round.

AI Science

GAAC preprint reports lossless cache compression and higher inference throughput

Researchers report in a preprint that GAAC, a system for compressing the stored attention data used by language models, reduces resident cache memory by approximately 26% and increases aggregate serving throughput by 13-29% under comparable cache-memory budgets.

Science AI

Preprint reports up to 32% decode energy savings from GPU clock control

A preprint reports that locking a GPU's processing clock recovered up to 32% of the energy used to generate language-model output, with minimal throughput loss. Researchers tested four attention architectures on an NVIDIA H200 using vLLM eager execution.

Infrastructure AI

Nvidia in talks to guarantee Nscale debt for West Virginia AI campus

Nvidia is discussing guarantees for billions in debt financing for Nscale's planned Anthropic data center in West Virginia, TechCrunch reported, citing The Information and Bloomberg.

AI Infrastructure

Five technology giants commit more than $3T to future spending

Alphabet, Meta, Microsoft, Amazon and Oracle have committed more than $3T to future spending on data centres, leases, equipment and energy, The Next Web reported, citing Bloomberg. Alphabet accounts for more than $1T and Meta for more than $800B.