We take cryptocurrency — so the coins you hold can buy real hardware.Pay in crypto, spend it on real hardware. Free insured shipping worldwide over $399. Plain, unbranded boxes, sent from Germany.Free insured shipping over $399. Your rate is locked for 30 minutes once checkout opens.Rate locked 30 minutes. Every item sealed, serial-checked and covered by a 24-month warranty.Sealed, 24-month warranty.

Saved items Sign in

ASUS

ASUS 90YT00A0-M0AA00 UGen300 USB AI Accelerator Hailo-10H 8GB LPDDR4 40 TOPS

SKU CH-C6KW-TQH9A MPN 90YT00A0-M0AA00

USB AI accelerator with Hailo-10H chipset delivering 40 TOPS at 2.5W, 8GB LPDDR4 memory, USB 3.1 Gen2 Type-C interface, and cross-platform support for local generative AI inference

  • TypeAI Accelerator
  • ColorSilver
  • SpecificationsChipset: Hailo-10H AI Performance: 40 TOPS @ INT4 | 20 TOPS

Option UGen300 USB AI Accelerator

  • Run classic and generative AI locally with the world's first USB edge accelerator
  • Deliver up to 40 TOPS at INT4 while drawing only 2.5 watts typical power
  • 8GB LPDDR4 onboard memory enables offline inference without cloud reliance
  • Plug-and-play USB 3.1 Gen2 Type-C at 10Gbps works on x86 and ARM
  • Access 150-plus pre-trained models via the online model zoo for LLM, VLM and vision tasks

Edge AI accelerator for developers and makers

The ASUS UGen300 is a USB AI accelerator built around the Hailo-10H chipset, delivering up to 40 TOPS at INT4 precision with 8GB of LPDDR4 memory onboard. It connects via USB 3.1 Gen2 Type-C at 10Gbps and runs on Windows, Linux and Android across x86 and ARM platforms. The device targets developers, makers, educators and industrial or SMB users who need local, offline inference for classic and generative AI workloads.

Thermals noise and power draw

Typical power consumption is 2.5 watts, allowing the accelerator to run from a standard USB-C port without an external power supply. The low draw keeps heat output minimal during sustained inference, and the fanless design means there is no audible noise in normal operation. This makes it suitable for quiet environments and space-constrained edge deployments where active cooling is impractical.

Bundled software and accessories

Access to an online model zoo provides 150 pre-trained models covering LLMs, VLMs, Whisper and vision networks, ready to deploy without training from scratch. Compatibility with TensorFlow, TensorFlow Lite, Keras, PyTorch and ONNX lets existing workflows run unchanged. Local execution keeps data on the device, improving privacy and removing cloud latency for real-time edge applications.

Highlights

  • Run classic and generative AI locally with the world's first USB edge accelerator
  • Deliver up to 40 TOPS at INT4 while drawing only 2.5 watts typical power
  • 8GB LPDDR4 onboard memory enables offline inference without cloud reliance
  • Plug-and-play USB 3.1 Gen2 Type-C at 10Gbps works on x86 and ARM
  • Access 150-plus pre-trained models via the online model zoo for LLM, VLM and vision tasks

Specifications

BrandASUS
ModelS-UGEN300-8GLD4-TC
TypeAI Accelerator
ColorSilver
SpecificationsChipset: Hailo-10H
AI Performance: 40 TOPS @ INT4 | 20 TOPS @ INT8
Interface: Compatible with USB3.1 Gen2 Type-C (10Gbps)
Memory: LPDDR 8GB
Memory Speed: 4266 MT/s
Power Consumption: 2.5 Watts (Typical)
Supported OS: Windows, Linux, Android
Supported Host Architectures: x86, ARM
Supported AI Framework: Keras, TensorFlow, TensorFlow Lite, PyTorch, ONNX
Supported AI Models: Vision Models (> 150+),
GenAI Models (LLM, VLM, Whisper)
Package Contents: 1 x Quick Start Guide
1 x Type-C to Type-C cable, white, 20 cm
Operating Temperature: 0°C ~ +40°C
Operating Humidity, RH: 20% ~ 85%, non-con
Shipping weight0.23 kg
Package size140 × 84 × 43 mm

Other versions of this model

1 more in the same range

Same family, different memory, cooler or clock — compare before you commit.

Questions about this item

How do I install the ASUS UGen300 and what step is often missed?

Plug the device into a USB 3.1 Gen2 Type-C port running at 10Gbps; the step often missed is confirming the port negotiates 10Gbps rather than falling back to a slower speed.

What maintenance or replacement will the UGen300 need later?

The 8GB LPDDR4 memory is soldered on board and the Hailo-10H chipset has no user-serviceable parts, so no field maintenance or component replacement is expected.

What limit appears first under sustained AI workloads?

The 2.5 watt typical power envelope caps sustained performance before thermal throttling becomes a factor.

New to this? Which coin should you pay with — what to check, what it costs to pay, and what is never asked for.

Also in this aisle

People compared these

Same shelf, same checkout — eight coins and a 30-minute rate lock.