Meta has come back to the consumer, and it has come back armed. Muse reads inboxes, drives a browser, books travel and pays. Meta has not led with intelligence. It has led with containment. A model that says the wrong thing is an embarrassment. An agent that does the wrong thing is now an incident.
OpenAI's Astra can develop zero-day exploits, while a US$1 billion defence push seeks to give critical infrastructure teams the same machine-speed advantage.
OpenAI’s Astra has reignited the AGI debate, with Jensen Huang declaring its arrival and researchers questioning the evidence. As autonomous AI advances, the race into 2027 will test who can turn greater capability into economic value while keeping human oversight firmly in place. The stakes climb.
Ayar Labs has secured $500 million in a Series E round to scale its co-packaged optics technology. Backed by NVIDIA and AMD, the company is replacing traditional copper interconnects with light-based data transmission to solve the growing power and bandwidth crisis in AI data centres.
Ayar Labs has closed a $500 million Series E funding round to accelerate the production of its optical I/O technology, which aims to solve the escalating power and data bottleneck crisis in large-scale AI systems.
Ayar Labs, a US-based semiconductor company founded in 2015, develops co-packaged optics (CPO) that use light to transfer data between chips instead of traditional copper wires. The company's core products, the TeraPHY optical I/O chiplet and SuperNova light source, are designed to be integrated directly into the packages of advanced AI processors and switches.
The Update: Ayar Labs has raised $500 million in a Series E round, bringing its total funding to $870 million and valuing the company at $3.75 billion. The round was led by Neuberger Berman and included new strategic investments from semiconductor giants AMD, Alchip, and MediaTek, alongside existing backer NVIDIA. Key institutional investors such as ARK Invest, Insight Partners, and the Qatar Investment Authority also participated.
"AI infrastructure is hitting a power wall driven by interconnect inefficiency. As bandwidth demands explode, copper becomes the bottleneck, consuming too much power and limiting AI throughput per watt and per dollar. Co-packaged optics overcomes these barriers, enabling thousands of GPUs to operate as a unified system." - Mark Wade, CEO and Co-founder, Ayar Labs
The capital will be used to scale high-volume manufacturing and testing of its co-packaged optics (CPO) solutions, which replace bandwidth-limited copper interconnects with high-speed, low-power optical links. The company plans to expand its global operations, including a new office in Hsinchu, Taiwan, to meet the demands of hyperscale AI customers. The technology is critical for enabling thousands of GPUs to operate as a single, unified system, overcoming the physical limitations of copper that create performance bottlenecks and drive up power consumption in modern data centres.
In an exclusive with @WSJ, we detailed how our CPO technology is solving one of the biggest hardware issues that’s causing bottlenecks in AI and revolutionizing the way chips are connected inside server racks.
— Ayar Labs (@AyarLabs) March 3, 2026
Read more:https://t.co/QrtHhqcdX6 pic.twitter.com/ecCY5Y5qMY
Why it Matters: The significant investment from major chipmakers like NVIDIA and AMD signals strong industry conviction that optical I/O is a necessary technology for the future of AI infrastructure. As AI models grow in complexity, the electrical resistance and power draw of copper wiring become a fundamental barrier to scaling performance efficiently.
Ayar Labs' approach of integrating light-based communication directly onto the chip package addresses this power wall head-on, promising substantial improvements in data throughput per watt. This shift is not just an incremental upgrade; it represents a foundational change in data centre architecture required to sustain the progress of AI.
Get the stories that matter to you.
Subscribe to Cyber News Centre and update your preferences to follow our Daily 4min Cyber Update, Innovative AI Startups, The AI Diplomat series, or the main Cyber News Centre newsletter — featuring in-depth analysis on major cyber incidents, tech breakthroughs, global policy, and AI developments.
NVIDIA’s US$96.2 billion quarter and 117% growth in data-centre revenue show AI compute becoming a global industrial market, while memory constraints and China uncertainty become harder commercial limits.
Brazil is investing US$444 million in AI infrastructure spanning Chinese and US technology. Its strategy avoids dependence on one power, using rival suppliers to build sovereign compute, domestic models and leverage, although genuine autonomy will depend on its energy, skills and strong execution.
Unitree’s Shanghai IPO puts a fast-growing humanoid-robot maker at the centre of China’s effort to turn embodied AI from a laboratory theme into an exportable industrial platform.
AMD's proposed Taalas acquisition targets the fast-growing AI inference market, adding specialised silicon technology designed around model-specific workloads to its Instinct, EPYC, Helios and ROCm platform.
Where cybersecurity meets innovation, the CNC team delivers AI and tech breakthroughs for our digital future. We analyze incidents, data, and insights to keep you informed, secure, and ahead. Sign up for free!
