Series CFuriosaAI

FuriosaAI designs AI inference chips that run LLMs at 2.25× GPU efficiency

Seoul AI-chip startup whose RNGD accelerator powers LG's EXAONE models with 2.25× better inference performance than GPUs — a Tensor Contraction Processor architecture now scaling to mass production.

شركة سيولية تصمم شرائح استدلال للذكاء الاصطناعي تشغّل نماذج إكساون من إل جي بكفاءة تفوق وحدات معالجة الرسومات 2.25 مرة — بمعمارية مخصصة تتوسع الآن نحو الإنتاج الضخم.

Next →
  1. The Bottleneck

    What was broken?

    Running large language models depends on expensive, power-hungry GPUs controlled by one vendor — datacenters and AI adopters need cheaper, more energy-efficient inference hardware.

  2. The Move

    Why it worked

    Built a purpose-built inference architecture (Tensor Contraction Processor) instead of repurposing graphics chips, then proved it where it counts: LG adopted RNGD for EXAONE after a two-year evaluation showing 2.25× better inference performance than competitive GPUs.

  3. The Trap

    Battle scars

    Nvidia's CUDA moat and massive scale; long, lumpy enterprise qualification cycles; and the capital intensity of taping out next-generation silicon while scaling RNGD production globally.

The problem

Every lab and enterprise running large language models rents the same scarce resource: power-hungry GPUs from a single dominant vendor. Inference — actually serving models to users — is where the power bill lands, yet purpose-built, energy-efficient alternatives were scarce, and buyers wanting supply-chain independence had few proven options.

How it works

FuriosaAI is a fabless chip company built around its Tensor Contraction Processor architecture, designed from scratch to minimize the data movement that wastes energy during inference. Its first chip, Warboy, targeted computer-vision workloads; its second chip, RNGD (Renegade), is optimized for LLMs and reasoning models. RNGD ships in servers that enterprises deploy to run LG AI Research’s EXAONE models across electronics, finance, telecommunications, and biotechnology applications.

Pain points

GPU scarcity and pricing power, datacenter power envelopes that cap how many models can be served, and governments and conglomerates hunting for sovereign, non-US chip supply.

Business model

Chip and system sales: RNGD accelerators and RNGD-powered servers sold to enterprises and AI labs, with LG AI Research as the anchor customer validating the hardware. A next-generation chip is in development to extend the roadmap.

Challenges

Dislodging Nvidia’s CUDA ecosystem moat, surviving long enterprise qualification cycles (LG evaluated RNGD for two years before adopting it), and funding next-generation tape-outs while scaling RNGD production to meet global demand.

Funding

  • Raised: $125M Series C bridge (Jul 2025, with Korea Development Bank, Industrial Bank of Korea, Kakao Investment, Keistone Partners, PI Partners participating), taking total funding to $246M (Verdict). Earlier the company was reportedly raising ~$48M (KRW 70B) in March 2025 (TechCrunch).
  • Valuation: $735M at the Series C bridge (Verdict).

Latest — July 2025

FuriosaAI closed the $125M Series C bridge to scale RNGD production globally and start its next-generation chip — weeks after announcing the LG AI Research partnership supplying RNGD to enterprises on the EXAONE platform, the payoff for staying independent after declining Meta’s reported $800M acquisition offer in March 2025.

Similar sparks

⚡️ LATE-STAGEVal: ~$2.3B

South Korea · 2026-09-25

Rebellions builds full-stack AI inference chips to break the GPU bottleneck

Fabless Korean chipmaker whose ATOM and REBEL NPUs plus RebelServer, RebelRack, and RebelPOD systems serve LLMs at a fraction of GPU power — Korea's first AI chip unicorn after absorbing SK Telecom's Sapeon.

بالعربي

شركة كورية تصمم شرائح استدلال للذكاء الاصطناعي مع منظومة برمجية وأنظمة مراكز بيانات تخدم النماذج اللغوية الكبيرة بجزء من استهلاك الطاقة — أول يونيكورن كوري في شرائح الذكاء الاصطناعي بعد اندماجها مع سابيون.

  • AI/semiconductors
  • #south-korea
  • #ai
⚡️ PRE-IPO$44.5M pre-IPO round (Mar 2021)

South Korea · 2026-09-16

Standigm designs new drug candidates end-to-end with workflow AI

Standigm's AI platform goes from disease target to first-in-class lead compound — mining biomedical data, designing novel molecules, and partnering with pharma to carry them toward the clinic.

بالعربي

تستخدم ستانديغم الذكاء الاصطناعي لاكتشاف أهداف دوائية جديدة وتصميم مركّبات رائدة من البداية حتى المرشّح الدوائي، بالشراكة مع شركات الأدوية العالمية.

  • biotech
  • #south-korea
  • #healthtech
⚡️ ACQUIREDKRW 267B (~$181M) Samsung stake increase to 35% (Dec 2024)

South Korea · 2025-05-21

Rainbow Robotics builds humanoids and cobots — now as Samsung's robot arm

KAIST-spinoff robot maker behind Hubo, Korea's first bipedal robot, plus collaborative arms and the wheeled RB-Y1 humanoid — a Samsung consolidated subsidiary since Samsung raised its stake to 35%.

بالعربي

شركة روبوتات انبثقت عن معهد كايست وصنعت هيوبو أول روبوت كوري يمشي على قدمين، وتنتج الأذرع التعاونية والروبوت شبه البشري آر بي-واي1 — وأصبحت شركة تابعة لسامسونغ بعد رفع حصتها إلى 35٪.

  • robotics/manufacturing
  • #south-korea
  • #ai
⚡️Spark playbook copied! Go build.