AI news, product launches and company developments.
OpenAI & AnthropicThe 2026 ICML conference highlights a strong focus on open frontier models and open AI infrastructure. These open models are becoming essential for advancing AI research. NVIDIA contributed significantly with 74 accepted papers, showcasing leadership in this area.
Countries are investing heavily in AI to boost their economies and enhance national security. AI is driving innovation in sectors like transportation, healthcare, and communications. This global push highlights AI's critical role in shaping future strategic priorities.
NVIDIA is enabling large-scale AI compute by partnering with capital investors to expand AI infrastructure. This approach supports the growing demand for continuous AI inference in production environments. The initiative aims to provide scalable, multi-tenant accelerated computing for efficient AI service delivery.
Amazon Bedrock now supports NVIDIA Nemotron and OpenAI GPT OSS models in AWS GovCloud (US). This enables US-based users to run large open-weight AI models with data residency compliance. The service offers multiple tiers and inference options for government and regulated industries.
Etched, a competitor to Nvidia, has reached a $5 billion valuation. The company has secured $1 billion in sales contracts for its AI inference chips. This shows strong market demand for alternatives to Nvidia's AI hardware.
NVIDIA has introduced the BioNeMo Agent Toolkit to accelerate AI applications in life sciences. This toolkit integrates with Claude Science, an AI workbench designed for scientific research. It enables researchers to run complex workflows faster using GPU-accelerated computing.
NVIDIA highlights how its inference software stack reduces token cost for AI workloads. The stack is optimized for GPUs, CPUs, and networking to deliver efficient performance. This approach helps organizations scale AI production with lower cost per token and energy use.
NVIDIA highlights three workflows to enhance Vision AI agent accuracy using synthetic data and fine-tuning within the Omniverse platform. These methods help convert real-world video data into actionable insights for industrial environments. The approach leverages OpenUSD and NVIDIA Omniverse to improve AI performance in practical applications.
Anthropic's Claude AI models are now running on NVIDIA GB300 Blackwell Ultra GPUs within Microsoft Azure. This integration is generally available through Microsoft Foundry, enabling enterprises to build autonomous and domain-specific AI agents. The collaboration enhances Azure-native AI capabilities with powerful computing resources.
Palantir has launched a new intelligent engine using NVIDIA Nemotron open models to support U.S. government agencies. This solution emphasizes secure AI deployment in closed environments while leveraging open source innovation. It highlights the ongoing importance of open models in advancing American AI capabilities.
Nvidia has long dominated the AI chip market, but companies like OpenAI, Google, Apple, and SpaceX are now developing their own custom chips. OpenAI revealed its Jalapeño inference chip made with Broadcom to reduce reliance on a single supplier. This shift aims to diversify hardware options and improve AI performance and control.
OpenAI announced its custom AI inference chip called Jalapeño, developed with Broadcom. This move aims to reduce reliance on Nvidia, which has dominated the AI chip market. Other major tech companies like Google and Apple are also creating their own chips to avoid single-supplier risks.
