Author HubMaximilian Schreiner
G42 is moving to lessen its dependence on Nvidia and is in talks with AMD, Cerebras Systems, and Qualcomm, according to a source familiar with the matter who spoke to Semafor. G42 also holds a stake in Cerebras Systems.
Of the five gigawatts planned for the UAE-US AI Campus, one gigawatt is already set aside for a Nvidia-powered Stargate data center. But these negotiations make it clear G42 doesn't want to rely solely on Nvidia. The company is aiming for a more diversified hardware base, partly in response to geopolitical tensions and concerns over supply chain dependencies.
OpenAI is planning a data center in India with at least 1 GW of capacity, Bloomberg reports. The company is searching for local partners, but details about the location and timeline are still unclear. OpenAI has already registered in India and aims to open a New Delhi office in 2025. The project could be part of the Stargate initiative, backed by SoftBank, Oracle, and OpenAI. In Norway, a Stargate center has been officially announced with an initial 230 MW, targeting 520 MW after expansion. Other Stargate projects include a planned 1 GW cluster in the United Arab Emirates, with the first 200 MW phase coming in 2026, and up to 4.5 GW of extra capacity in the US through Oracle.
X.AI has introduced Grok Code Fast 1, a new AI model designed specifically for agent-based programming. The company says it uses a "new architecture," was trained on real-world programming data, and is built to be fast and cost-effective. Grok Code Fast 1 supports a wide range of programming languages and is intended to handle tasks like bug fixes and project setup on its own. Early integration partners include GitHub Copilot, Cursor, and others, where the model is available for free testing for a limited time. Pricing is set at $0.20 per million input tokens and $1.50 per million output tokens. X.AI hasn't released benchmark comparisons, sharing only a single SWE-Bench score of 70.8 percent. According to initial user reports, the model is fast, but struggles with more complex tasks and makes frequent mistakes. Grok Code Fast 1 seems positioned as an alternative to smaller models like GPT-5-nano.
Microsoft's Copilot AI assistant is now available on select Samsung TVs and monitors from the 2025 lineup, including Neo QLED, OLED, The Frame, and the M7, M8, and M9 monitors. Copilot is integrated through the Tizen operating system and appears in the "Samsung Daily+" interface, where it can be launched using the microphone button on the remote. Users can ask questions, get recommendations, or look up information about movies and TV shows. Answers are delivered as spoken responses and visual cards with images, ratings, and additional details. An animated character on the screen matches the conversation with facial expressions and lip syncing. Connecting a Microsoft account unlocks personalized suggestions. Copilot is free to use but is initially limited to select regions.
Google will invest an additional $9 billion in expanding its cloud and AI infrastructure in Virginia by 2026. The plan includes building a new data center in Chesterfield County and expanding existing facilities in Loudoun and Prince William counties, according to a company blog post. This move is part of a larger trend, with tech giants pouring money into data centers to keep up with rising demand for AI computing. Bloomberg reports that Google's annual investment in this area now stands at $85 billion - $10 billion more than previously planned. A spokesperson noted that data centers in Virginia often take years to come online due to power supply challenges. Google has not announced a completion date for the new Chesterfield facility but expects construction to take 18 to 24 months. Dominion Energy will provide electricity for the project.