Name: Furiosa RNGD
Brand: FuriosaAI
Availability: InStock

Eight red GPU units mounted inside a server chassis with cables and circuit boards visible.

"For the launch of our new Seoul office, we showcased our gpt-oss 120B model running live on Furiosa’s flagship RNGD accelerator marking the first time our models have run on Korean silicon.

This was a huge milestone, showing what’s possible when world-class AI meets world-class hardware innovation."

Danial Mirza, Applied AI

Proven in Production

Output throughput (tokens/s)

Performance comparison chart showing Furiosa RNGD ×4 versus RTX PRO 6000 ×4 across multiple batch sizes. The graph plots interactivity (TPS per user) against aggregate throughput, illustrating that RNGD delivers higher throughput while maintaining target service-level objectives, outperforming the comparison system at batch sizes 16, 32, and 64.

Concurrency /kW

Bar chart comparing TPS per user SLO of 20: red bar 63 versus gray bar 35.6 with 1.8x difference.

Bar chart comparing TPS/user SLO of 31.6 versus 16.7 with a 1.9 times increase noted.

Bar chart showing Furiosa TPS per user is 15.3, double the 7.5 TPS of the other measure.

RNGD ×4 (Mar 13)

RTX PRO 6000 ×4 (vLLM 0.18.0)

Qwen 3-32B, FP8, 1K / 1K

THE RENEGADE AI ACCELERATOR

Composite product image showcasing Furiosa RNGD AI accelerator hardware. The image features the external accelerator card with a red thermal enclosure alongside the exposed internal PCB, revealing the RNGD processor, power delivery components, memory interfaces, and board-level architecture. The side-by-side presentation highlights both the industrial design and engineering implementation of the high-performance AI accelerator.

RNGD (pronounced “renegade”) delivers high-performance LLM and multimodal deployment capabilities while maintaining a radically efficient 180W power profile.

512TFLOPS

64TFLOPS (FP8) x 8 processing elements

48GB

HBM3 memory capacity

2 x HBM3

CoWoS-S, 6.0Gbps

256MB SRAM

384TB/s on-chip bandwidth

1.5TB/s

HBM3 memory bandwidth

180W TDP

Targeting air-cooled data centers

PCIe P2P support for LLMBF16, FP8, INT8, INT4 support

Multiple-Instance and VirtualizationSecure boot & model encryption

Start testing with Furiosa Access

Data center deployment of a Furiosa RNGD AI inference server with multiple accelerator cards installed in a rack-mounted system, delivering scalable performance, high density, and energy-efficient execution for large-scale AI and LLM workloads.

Interested in evaluating RNGD? The Furiosa Access Program provides a structured path for customers and partners to evaluate, integrate, qualify, and deploy Furiosa accelerators through both online and offline access. Available worldwide.

Furiosa Access Locations

Seoul

KOR

Bay Area

USA

Lisbon

PRT

Johor Bahru

MYS