← sources

FURIOSA IR snapshot

source: ir (ir:FURIOSA)published 2026-09-04 02:34ingested 2026-09-04 02:51

extract

Renegade 2026 Keynote The full keynote sessions are now available. 3 KW INFERENCE APPLIANCE FOR AGENTIC SYSTEMS RNGD enables 4x more inference capacity Tensor contraction, not matmul The fundamental computation of modern day deep learning is tensor contraction, a higher dimensional generalization of matrix multiplication. However, most commercial deep learning accelerators today incorporate fixed-sized matmul instructions as primitives. RNGD breaks away from that, unlocking powerful performance and efficiency. INFERENCE WITHOUT CONSTRAINTS Performance Efficiency Programmability SOFTWARE FOR LLM DEPLOYMENT Start testing with Furiosa Access Blog FuriosaAI at ICML 2026: Advancing full-stack software efficiency White Paper: Benchmarking RNGD on Backend.AI for 1.3–1.5x greater efficiency in enterprise AI inference FuriosaAI partners with I/ONX and Velox for new 15 MW AI data center in Sweden

retained text

Furiosa RNGD accelerates AI transformation
FuriosaAI closes $620 million Series D
[FuriosaAI closes $620 million Series D](/blog/furiosaai-620-million-series-d)
FuriosaAI expands European AI infrastructure with RNGD deployment at Equinix’s Lisbon data center
Furiosa SDK 2026.3: A new kernel framework, and the models it unlocks
FuriosaAI and Broadcom partner on next-gen inference for Agentic AI
RNGD hardware, built for AI transformation at any scale.
Furiosa NXT RNGD Server
[Learn more about the Furiosa RNGD PCIe CardFuriosa NXT RNGD Server](/nxt-rngd-server)
Furiosa RNGD PCIe Card
[Learn more about the Furiosa RNGD PCIe Card](/rngd)
Drop-in GPU replacement. Migrate your entire client code today.
pip install furiosa-llm
furiosa-llm serve [model]
# Serving on :8000 (OpenAI-compatible)from openai import OpenAI
client = OpenAI(
    base_url="http://localhost:8000/v1"
)
client.chat.completions.create(model="[model]", ...)from furiosa_llm import LLM
llm = LLM(
    model="[model]",
    dtype="[precision]",
)
out = llm.generate(prompts)
Start Your RNGD Evaluation
Bare metal
API endpoint
Blog
FuriosaAI closes $620 million Series D to scale efficient, production-ready AI inference
Furiosa SDK 2026.4: Enhanced MoE performance, hierarchical caching, expanded model coverage
FuriosaAI establishes Singapore hub to drive APAC expansion and accelerate global RNGD deployment

document fetched 2026-11-12 03:00 · canonical https://furiosa.ai