Intel Unveils Upcoming AI Chip to Challenge Nvidia and AMD
Intel plans to launch a new AI chip by the end of the year, aiming to offer lower costs and improved cooling compared to rival products from Nvidia and AMD. The 'Crescent Island' GPU is designed for AI inference tasks, an area of increasing demand in data center and cloud computing applications.
Nvidia’s Vera Chip Targets $200 Billion AI Inference Market
Nvidia has unveiled its new Vera central processor, aiming to capture a $200 billion market focused on AI inference workloads. CEO Jensen Huang highlighted the Vera chip as a critical component in Nvidia’s strategy to maintain its AI hardware leadership as major cloud providers and chip competitors ramp up their own silicon offerings. Despite record earnings, the company faces supply constraints and intensified competition in the AI hardware sector.
UK AI Chip Startup Fractile Secures $220M to Boost Inference Performance
Fractile, a UK-based AI chip company, has raised $220 million in Series B funding to develop specialized hardware aimed at improving AI inference speed and efficiency. The investment highlights mounting industry focus on overcoming inference bottlenecks in advanced AI applications.
Nebius Acquires Eigen AI for $643M to Boost AI Inference Capabilities
Nebius, an Amsterdam-based AI infrastructure company, has acquired San Francisco’s Eigen AI in a $643 million deal. The acquisition expands Nebius's AI inference capabilities and brings new engineering talent to its portfolio.
Samsung Begins Mass Production of Nvidia Groq 3 LPU on 4nm Process
Samsung Electronics has started large-scale manufacturing of Nvidia's Groq 3 Language Processing Unit using its 4-nanometer process, marking a key development for both companies in the artificial intelligence hardware sector. The move underscores Nvidia’s ongoing strategy to broaden its AI hardware beyond traditional GPUs, while positioning Samsung as a major foundry partner in the global AI chip race.
NVIDIA Blackwell Ultra Offers Major Efficiency Gains for Agentic AI
NVIDIA’s Blackwell Ultra GPUs deliver up to 50 times higher efficiency for agentic AI and coding assistant workloads, according to new benchmark results. The platform significantly reduces operational costs for cloud providers and enables wider deployment of advanced language models. Cloud infrastructure firms including CoreWeave, Microsoft, and Oracle have begun large-scale deployments using this technology.