Search AI Products and News

Google recently released a new version of its Gemma3 series, exciting many AI enthusiasts. Just a month after its initial launch, Google released a Quantization Aware Training (QAT) optimized version of Gemma3, aiming to significantly reduce memory requirements while maintaining model quality. Specifically, the QAT-optimized Gemma3 27B model reduces VRAM requirements from 54GB to 14.1GB, meaning users can now run it on a single NVIDIA RTX 3090.

Nvidia recently announced plans to establish over one million square feet of manufacturing space in Arizona and Texas for the production and testing of AI chips. This represents a significant move by Nvidia to bring some of its manufacturing back to the United States. Nvidia's Blackwell chip is already in production at TSMC's Arizona facility. Additionally, Nvidia is establishing "supercomputer" manufacturing plants in Texas, partnering with Foxconn in Houston and Wistron in Dallas. In Arizona, Nvidia is collaborating with Amkor and others.

AI chip giant Nvidia announced it will partner with manufacturing collaborators to design and build its first AI supercomputer on US soil, marking a significant step in its supply chain strategy. Nvidia has commissioned over one million square feet of manufacturing space for production and testing of its latest Blackwell AI chips in Arizona, and for the manufacturing and testing of the AI supercomputer in Texas. Nvidia's ecosystem partners are expected to invest $500 million to support the construction of this AI infrastructure. While the...

Quantum artificial intelligence startup SandboxAQ announced the successful completion of its Series E funding round, raising $450 million. This round attracted investment from industry giants including Google, Nvidia, and BNP Paribas, bringing SandboxAQ's total funding to $950 million. The company stated that the funds will be used to accelerate the development of its large quantum models and foster collaborations across various industries. Image note: Image generated by AI, image licensing provider Midjourney.

NVIDIA, a global leader in chip and AI technology, recently launched a groundbreaking new open-source large language model, Llama 3.1 Nemotron Ultra 253B, generating significant excitement within the AI community. Built upon Meta's Llama-3.1-405B, this model boasts innovative optimizations that surpass competitors like Llama 4 Behemoth and Maverick in performance, while demonstrating superior resource efficiency and exceptional multi-tasking capabilities.

On April 8th, 2025, NVIDIA launched Llama 3.1 Nemotron Ultra 253B, an open-source model optimized from Llama-3.1-405B. With 25.3 billion parameters, it surpasses Meta's Llama 4 Behemoth and Maverick, becoming a focal point in the AI field. This model demonstrates superior performance in benchmarks such as GPQA-Diamond, AIME 2024/25, and LiveCodeBench, achieving inference throughput comparable to DeepSeek.

According to The Information, Nvidia is on the verge of acquiring Lepton AI, a prominent AI startup, in a deal reportedly worth hundreds of millions of dollars. This move signals Nvidia's aggressive push into the cloud computing and enterprise software market, aiming to compete with major cloud providers like Amazon and Google. Founded two years ago by renowned scientist Yangqing Jia and his team, Lepton AI focuses on building new infrastructure for the AI era, simplifying the building and deployment of AI models. Unlike many startups...

Reports suggest that Nvidia, the global semiconductor giant, is nearing a deal to acquire the startup Lepton AI for potentially hundreds of millions of dollars. Founded two years ago, Lepton AI focuses on leasing servers equipped with Nvidia AI chips to other businesses. According to The Information, the deal is progressing rapidly, although Nvidia has yet to officially comment.

Elon Musk's xAI is collaborating with Nvidia, Microsoft, and BlackRock to significantly increase investment in artificial intelligence infrastructure. This partnership aims to accelerate advancements and development in the field of AI.

Former Intel CEO Pat Gelsinger recently criticized Nvidia's pricing strategy for its AI GPUs on the Acquired podcast during Nvidia's 2025 GPU Technology Conference, arguing that the high cost makes them unsuitable for large-scale AI inference tasks. Gelsinger highlighted that inference is crucial for deploying AI models and that the industry should focus more on inference, a domain where Nvidia's technology falls short in terms of cost-effectiveness. Image source omitted.