American multinational technology company. Ultrathink has published 118 stories on NVIDIA. Latest, on : FuriosaAI Raises $620M at $2.9B as Korea's State Fund Backs Its RNGD Inference Chip.
FuriosaAI closed a $620M Series D at about $2.9B post-money, with $280M from Korea's National Growth Fund. It plans 20,000 RNGD inference cards in 2026, 40,000+ in 2027, and a 2nm Gen 3 chip with Broadcom.
SWI Compute, part of Euronext-listed SWI Group, launched with 3.6 GW of power, a ~$1B NVIDIA order and a 30 MW UK AI factory of ~7,000 Rubin GPUs due by end of 2027, aiming for 2.3 GW by 2029.
Google Cloud A5X will pair NVIDIA Vera Rubin systems with Google's network and cooling, with up to 960,000 GPUs planned for AI workloads at global scale.
Dream Labs launched with a $90M seed led by B Capital and Kindred Ventures and a public API for Dream-1, a robot foundation model priced at $3/hour that controls unseen two-armed robots zero-shot.
NVIDIA RTX Spark laptops pair Blackwell graphics, Grace CPUs and up to 128GB of unified memory. Here is what that means for practical local AI on Windows.
NVIDIA values a five-year UK agentic safety compute donation at $250M ($50M/year). Majority to the wider research community; proportion to Oxford BOLD, UCL SOFAIR, Imperial. UK-hosted via Cloud Partners TBD.
Nous Research confirmed a $90M Series B at a $1.5B valuation led by Robot Ventures, per TechCrunch, bringing total funding to $158M. Hermes Agent has 24M+ clones by Nous’s estimate.
GMI Cloud raised $668M — $223M Series B led by ARCHIV with NVIDIA in, plus a $445M CTBC credit facility — to expand GPU capacity across the U.S., Taiwan and APAC. Contracted ARR is company-reported at 9× YE2025.
Reactor added NVIDIA and Sapphire Ventures to its Lightspeed-led Series A, bringing total funding to $74M for real-time world-model inference infrastructure. Valuation and October tranche size undisclosed.
Boost Run (BRUN) signed a ~$525.6M five-year dedicated GPU cloud deal with Cohere for NVIDIA GB300 NVL72 capacity; acceptance expected early Q2 2027; TCV now over $2.6B.
Iterate.ai generally available Lifeboat claims 2–6× concurrent agent sessions per GPU with confidential computing; company benches show 2,048 sessions on one RTX PRO 6000 Blackwell.
SharonAI (NASDAQ: SHAZ) closed a US$356M senior secured GPU-backed SPV facility at 9.95% fixed to fund contracted AI Factory deployments toward 68,000+ NVIDIA GPUs by mid-2027.
Samsung Electronics and five affiliates are investing a combined $1 billion in Helix, KKR's AI infrastructure company led by ex-AWS CEO Adam Selipsky.
Nscale Limited announced a $3.36 billion pre-IPO convertible raise led by Third Point on Sept. 25, 2026. Company-reported TCV is over $103 billion.
Basecamp Research announced a $140 million Series C led by S32 on Sept. 23, 2026 to train new EDEN models for therapeutics. No valuation on the primary wire.
Verda raised $189 million in an oversubscribed Series B led by Emergence Capital. The Helsinki AI cloud, also known as DataCrunch, calls itself Europe's latest unicorn.
Crusoe announced the initial close of an anticipated $3.9 billion Series F at a $30.9 billion post-money valuation. Co-led by Atreides, Mubadala, and Valor.
At Dreamforce 2026 in San Francisco, Marc Benioff called SaaSpocalypse talk crazy nonsense and said AI may end software that makes humans do the work.
NVIDIA agreed to acquire Hugging Face for $12,930,300,000. The 8-K splits ~$11.9B to stockholders plus up to ~$1.0B retention. Close is expected in 1H 2027.
Sept 2 in Austin: official agenda puts AWS energy and OpenAI data-center ops on stage at 10:00 CT, then CoreWeave. No public floor quotes yet.
NVIDIA sovereign AI Indonesia initiative with Indosat and UGM bypasses traditional cloud giants. Learn how this project secures Southeast Asian tech.
Together AI Chennai unified cluster deploys 10,000 NVIDIA B300 GPUs to bypass networking bottlenecks. Learn why this architectural shift matters.
Alibaba's Qwen team has quietly uploaded Qwen3.8-2.4T-A95B-FP8 to Hugging Face. Here is what this highly efficient MoE model means for open-source AI.
Alibaba leaks Qwen3.8-2.4T-A95B-FP8 on Hugging Face, signaling a massive 2.4 trillion parameter MoE model to rival Meta's Llama open-source dominance.
Learn how the new 800 VDC power architecture co-developed by NVIDIA, Google, and Microsoft bypasses legacy AC-to-DC conversion losses.
NVIDIA Nemotron 3.5 Lightning and NeMo Switchyard debut to solve the cost and latency bottlenecks of multi-agent AI workloads. Here is our technical analysis.
The vLLM inference engine v0.27.0 update introduces breaking PyTorch 2.13 changes while optimizing for Nvidia Blackwell GPUs and Moonshot AI's Kimi K3.
The Meta Muse Glimmer release delivers a 30B local AI model that outperforms Google's Gemma4-31B, running under 20GB VRAM on consumer GPUs.
Nvidia is coordinating a historic $500 billion AI infrastructure funding package, shifting massive data center CapEx directly to Wall Street.
Meta Muse Glimmer has launched. Discover how this new open-source, local, agentic, multimodal AI framework runs directly on standard consumer hardware.
The Firebird Sovereign AI factory in Armenia, built with Dell and Nvidia, secures 70k Blackwell and Rubin GPUs to challenge US tech dominance. Read more here.
Samsung next-generation memory architectures use advanced wafer-bonding to solve critical thermal and density bottlenecks in modern AI data centers.
Samsung's new zHBM AI memory technology uses advanced wafer bonding to bypass the data center memory wall and outpace traditional HBM4 architectures.
Liquid AI LFM2.5 delivers a 128K context and local tool-calling, proving on-device AI models can run agentic workloads without costly cloud APIs.
AMD acquires Taalas to hardwire AI models directly into silicon, bypassing costly GPUs to solve the industry's massive inference scale and power bottlenecks.
The AI data center bottleneck is tightening as local opposition and grid constraints block new construction, stalling the compute plans of tech giants.
Hugging Face inference providers now include Baseten, allowing developers to deploy open-source models instantly to production-grade serverless compute.
Enterprise AI leader Omilia raises $67M Series B funding. Discover how its capital-efficient, $60M ARR model beats cash-burning foundation startups.
Anthropic is building custom AI chips to optimize Claude and escape Nvidia's premium pricing, shaking up its partnerships with Google and AWS.
The Sarvam AI Nvidia investment marks a geopolitical shift, valuing the startup at $1.51B as India builds its sovereign trillion-parameter AI model.
The global venture funding surge in July 2026 reached a record $65 billion, driven by massive sovereign-scale AI bets and multi-billion-dollar rounds.
Discover how a space-based AI supercomputer from Nvidia and SpaceX bypasses terrestrial grid and cooling limits using Space-1 Vera Rubin modules.
NVIDIA AI storage gets a massive upgrade with the Vera CPU and BlueField-4 STX platform, eliminating x86 bottlenecks for agentic AI workloads.
Discover how the Nvidia-backed Open Secure AI Alliance is establishing AI agent security standards to defend against autonomous threats and secure enterprise deployment.
The new Mixture-of-Kittens MoE megakernel by Cursor bypasses Blackwell Grace CPU bottlenecks, boosting MXFP8 throughput by 2.37x on NVL72 hardware.
NVIDIA joins the NSF AI Hubs program, embedding its hardware and software stack into US academic research to secure long-term market dominance.
The Texas data center halt exposes severe ERCOT grid vulnerabilities, creating a massive bottleneck for AI compute and forcing tech giants to pivot.
Discover how the Nvidia Alpamayo 2 Super open reasoning model disrupts autonomous driving by offering a powerful open-source alternative to closed ecosystems.
London-based OLIX AI chip startup reaches a $3.3B valuation in a $312M Series B to tackle the data center interconnect bottleneck with silicon photonics.
NVIDIA B300 Blackwell Ultra GPUs solve the memory bottleneck for reasoning models like DeepSeek-R1. Discover the specs of the new 144 PFLOPS DGX system.
The Naver AI Factory project secures $10B from NVIDIA and Brookfield to build massive sovereign AI infrastructure without wrecking Naver's balance sheet.
The new Nvidia Vera CPU uses custom Olympus cores to accelerate agentic AI workloads, locking cloud providers into Nvidia's complete data center stack.
Apple CEO Tim Cook hints at a Siri AI subscription via iCloud+ to offset massive LLM compute costs. Here is what it means for consumer AI economics.
An unprecedented AI stock rout has wiped out 67% of the 'Situational Awareness' index in July 2026, signaling a harsh cooling of the AI investment bubble.
Citadel buys Situational Awareness's stock portfolio after massive AI losses. Read how Ken Griffin is capitalizing on the AI market correction.
Nscale acquires Anyscale, creator of Ray, to merge GPU infrastructure with software orchestration and challenge big tech hyperscalers.
The SK Hynix HBM sell-off and 16% Korean stock plunge expose growing market anxiety over whether the AI hardware infrastructure boom is starting to cool.
A sudden 16% crash in Korean equities highlights growing anxiety over SK Hynix AI chip demand and the sustainability of the generative AI hardware boom.
The deepening AI chip stock sell-off reflects growing Wall Street anxiety over the massive capital expenditure required for generative AI infrastructure.
Discover how agentic AI in scientific computing is redefining research as OpenAI maps out a shift from static simulations to autonomous discovery agents.