October 1, 2025

analysis  
Meta Acquires Rivos for $2B, Intel & AMD Alternatives

The artificial intelligence landscape continues to evolve rapidly, with major tech players making strategic moves. Meta is reportedly acquiring AI chip startup Rivos for approximately $2 billion. This acquisition aims to bolster Meta's in-house AI hardware development, particularly its custom AI accelerators known as MTIA, and reduce its reliance on Nvidia. Rivos specializes in designing chips based on the open-source RISC-V architecture, an alternative to established designs from companies like Arm, Intel, and AMD. Meta plans to leverage this acquisition to gain more control over its AI infrastructure and potentially lower costs, with a goal of using its own chips for AI training by 2026. Meanwhile, Nvidia CEO Jensen Huang dismisses Wall Street's predictions of a growth slowdown, citing immense and expanding demand for AI compute power. He also defended Nvidia's significant investment in OpenAI, viewing it as a strategic move for a future trillion-dollar company. In the realm of AI applications, Google's head of Search, Liz Reid, asserts that AI is augmenting, not replacing, search functions, enhancing users' ability to ask complex questions while traditional search results remain vital. Google has also released Tunix, a new JAX library designed to simplify the fine-tuning of Large Language Models. Microsoft is addressing the challenge of AI data center overheating with a new liquid-cooling technology that pumps coolant directly through etched silicon chips, offering a more effective solution than traditional methods. On the educational front, Will.i.am is set to teach a new AI course at Arizona State University, focusing on personalized AI agents. Imagine Learning is launching an AI course for middle and high school students aimed at developing critical judgment and ethical reasoning about AI, scheduled for Back to School 2026. Georgia is investing $65 million in a statewide initiative, Georgia AIM, to train its workforce in AI and automation for manufacturing jobs. San Francisco is piloting an AI chatbot to assist RV dwellers navigating a new parking ban, marking the city's first use of AI for direct resident interaction. Finally, a Goldman Sachs strategist advises investors to focus on how companies are effectively utilizing AI, rather than on potential regulatory hurdles, emphasizing the importance of AI implementation in third-quarter earnings reports.

Key Takeaways

  • Meta is reportedly acquiring AI chip startup Rivos for around $2 billion to enhance its in-house AI hardware development and reduce dependence on Nvidia.
  • Rivos designs AI chips using the RISC-V architecture, an open-source alternative to designs from Arm, Intel, and AMD.
  • Nvidia CEO Jensen Huang believes AI demand will continue to grow significantly, dismissing forecasts of a slowdown and defending Nvidia's investment in OpenAI.
  • Google's head of Search states that AI is augmenting search capabilities, not replacing them, while traditional search results remain important.
  • Google has released Tunix, a new JAX library to simplify the fine-tuning of Large Language Models.
  • Microsoft has developed a liquid-cooling technology for data centers that circulates coolant directly through silicon chips to prevent AI hardware from overheating.
  • Will.i.am will teach a new AI course at Arizona State University focusing on personalized AI agents.
  • Imagine Learning will launch an AI course for middle and high school students in Back to School 2026, focusing on critical judgment and ethics in AI.
  • Georgia is launching a $65 million initiative, Georgia AIM, to train its workforce in AI and automation for manufacturing.
  • San Francisco is testing an AI chatbot to provide information to RV dwellers regarding a new parking ban.

Meta buys AI chip startup Rivos to boost its own tech

Meta is acquiring the chip startup Rivos to improve its in-house AI hardware and reduce costs. Rivos designs chips using the RISC-V architecture, an alternative to established brands. This move could help Meta gain more control over its AI technology and potentially lower expenses associated with relying on outside chip suppliers. The acquisition highlights a trend among big tech companies to develop more of their hardware internally.

Meta's $2B bet on Rivos aims to challenge Nvidia's AI chip dominance

Meta is reportedly buying Rivos, an AI chip startup, to strengthen its custom semiconductor development and lessen its dependence on Nvidia. Rivos is creating RISC-V based AI chips, an open-source alternative that avoids costly licenses. Meta aims to use its own chips for AI training by 2026, and acquiring Rivos would speed up this plan by adding talent and technology. The deal, valued around $2 billion, is part of Meta's strategy to control key AI infrastructure and cut costs.

Meta reportedly pursuing Rivos AI chip startup acquisition

Meta Platforms is reportedly in talks to acquire Rivos, a startup based in Santa Clara that focuses on creating chips for AI data centers. This acquisition would help Meta enhance its internal AI hardware capabilities. The deal details have not yet been made public.

Meta buys Rivos to boost AI chip development and reduce Nvidia reliance

Meta is set to acquire Rivos, a startup specializing in RISC-V AI chips, to strengthen its internal chip development and decrease reliance on Nvidia. Rivos designs GPUs and AI accelerators using the open-source RISC-V architecture. Meta has been developing its own AI accelerators, the MTIA chip, and acquiring Rivos could accelerate this effort. The deal, with Rivos valued around $2 billion, aims to give Meta more control over its AI hardware.

Meta plans Rivos acquisition to advance its AI chip efforts

Meta intends to acquire the chip startup Rivos to enhance its in-house semiconductor development. Rivos, based in Santa Clara, California, focuses on designing chips using the RISC-V architecture, an open-source alternative to existing chip designs. This acquisition aims to accelerate Meta's work on its custom AI chip accelerators, known as MTIA. Meta has been a significant customer of Rivos and sees this move as a way to speed up its custom silicon development.

Meta to acquire Rivos to boost in-house AI chip development

Meta plans to acquire chip startup Rivos to strengthen its in-house semiconductor efforts. Rivos, based in Santa Clara, California, designs chips using the RISC-V architecture, an open-source alternative to those from Arm, Intel, and AMD. Meta aims to expand its work on custom-built AI chip accelerators called MTIA. The company has been a major customer of Rivos and believes this acquisition will accelerate its custom silicon development.

Rapper Will.i.am to teach AI class at Arizona State University

Musician and tech innovator Will.i.am is joining the faculty at Arizona State University (ASU) to teach a new course called 'The Agentic Self.' This class will focus on artificial intelligence and how to build personalized AI agents. The course will be taught from Will.i.am's studio and will be available to students online and in person. This collaboration is part of ASU's effort to prepare students for the evolving job market influenced by AI.

Imagine Learning launches new AI course for students

Imagine Learning has announced 'The AI Question,' the first course in a new AI learning pathway for middle and high school students. Launching for Back to School 2026, the course aims to help students develop critical judgment and ethical reasoning regarding AI. It will explore how AI works, analyze bias in machine learning, and examine the future of work. This initiative is designed to equip students with essential skills for an AI-driven world.

San Francisco tests AI chatbot for RV dwellers facing parking ban

The San Francisco Mayor's Office of Innovation is testing an AI chatbot to help RV dwellers understand the city's new ban on long-term street parking. The chatbot will answer questions about the enforcement rollout and a temporary permit system for housing assistance. This tool aims to provide accurate and timely information to marginally housed residents as the ban takes effect on November 1st. This marks the city's first use of an AI tool to directly interact with residents.

Georgia initiative aims to prepare workforce for AI and automation

Georgia is launching a statewide initiative called Georgia AIM (Artificial Intelligence in Manufacturing) as part of a $65 million federal effort to train the future workforce. The program focuses on teaching students how to use and control technology like AI, robotics, and drones, rather than be replaced by them. With a significant portion of manufacturing jobs already using automation, the state aims to fill millions of potential tech-enabled jobs by ensuring students are trained in these advanced skills.

Nvidia CEO Jensen Huang dismisses Wall Street AI forecasts

Nvidia CEO Jensen Huang is pushing back against Wall Street analysts who predict a slowdown in the company's growth after 2027. Huang believes these forecasts underestimate the massive demand for AI, driven by increasing user engagement and the growing compute power needed per user. He also defended Nvidia's $100 billion investment in OpenAI, calling it a strategic move to support a future trillion-dollar company. Huang stated that Nvidia's opportunity in AI is much larger than current consensus estimates suggest.

Goldman Sachs strategist advises investors to focus on AI use

David Kostin, Goldman Sachs' Chief US Equity Strategist, suggests that investors should concentrate on how companies are utilizing artificial intelligence rather than worrying about potential government shutdowns. He emphasizes that third-quarter earnings reports will be crucial for understanding AI's impact on businesses. Kostin believes that a company's strategic implementation of AI will be a key indicator of its future success.

Google's head of Search says AI augments, not replaces, search

Liz Reid, Google's head of Search, stated that artificial intelligence is enhancing search capabilities rather than replacing them. She believes AI acts as a layer that helps reinvent search, allowing users to ask complex questions effortlessly. Reid emphasized that traditional search results, or 'blue links,' will continue to be important, especially for content creators offering rich and trusted information. The use of AI is leading to longer, more contextual queries from users.

Google releases Tunix, a new JAX library for AI model training

Google has introduced Tunix, a new open-source, JAX-native library designed to simplify the process of aligning Large Language Models (LLMs) after initial training. Tunix offers a suite of algorithms for supervised fine-tuning, preference tuning, and reinforcement learning, optimized for performance on TPUs. The library provides a 'white-box' design for easy customization and seamless integration with the JAX ecosystem. Tunix aims to make it easier for developers to create production-ready AI models.

Microsoft develops liquid-cooled chips to prevent AI data center overheating

Microsoft engineers have created a new liquid-cooling technology using microfluidics to manage heat in data centers. This method pumps coolant through tiny channels directly etched into silicon chips, making it up to three times more effective than traditional cooling. The technology aims to prevent AI hardware from overheating, which could limit data center performance in the coming years. Microsoft hopes this innovation will allow for more intensive computational workloads and potentially enable safe overclocking of chips.

Developers must ensure AI-generated code is secure

The increasing use of AI in software development raises cybersecurity concerns, as developers integrate AI-generated code into their projects. While AI boosts productivity, it also escalates complexity and the risk of AI-powered attacks. Developers must be accountable for the security of AI-assisted code by understanding fundamentals, validating outputs, and staying adaptable. Establishing clear organizational policies and prioritizing security from the start are crucial for building resilient software in the age of AI.

Sources

AI chips Meta Rivos RISC-V Nvidia semiconductors AI hardware custom silicon AI training AI data centers AI development AI technology AI workforce AI education AI agents AI learning AI chatbot AI search AI models LLMs AI code AI security liquid cooling Google JAX Tunix Microsoft Goldman Sachs investors AI use Georgia AIM automation robotics drones Will.i.am Arizona State University Imagine Learning San Francisco RV dwellers Nvidia CEO Jensen Huang AI forecasts AI demand OpenAI