Nvidia Locks Down SK Hynix Memory Supply in Landmark 00 Billion AI Deal
Nvidia and SK Group signed a 00 billion+ partnership to secure AI memory supply and build massive AI data centers.
Technology News
Latest articles in Technology
Global Chip Race Fractures Into Four Strategies as US Advanced Packaging Gap Persists After $265B Bet
The global semiconductor race has split into four distinct national strategies. The US, despite $265 billion in investments, still cannot finish its own AI chips at home — every accelerator built at TSMC’s Arizona fab must travel to Taiwan for CoWoS advanced packaging until at least 2028.
New Programmable Photonic Chip Can Slow Light on Command, Paving Way for Faster Optical Computers
Scientists at Seoul National University have developed a programmable photonic integrated circuit that can slow light on demand, enabling faster and more energy-efficient optical computing.
OpenAI makes chatgpt health widely available with medical record integration
OpenAI launched Health in ChatGPT on July 23, making it broadly available to all US users. The feature allows users to securely connect medical records through a partnership with b.well and sync wellness apps like Apple Health. OpenAI says the new GPT-5.6 Sol model is its 'strongest yet for health,' capable of explaining lab results and tracking health patterns.
OpenAI says its AI models went rogue during security test and hacked Hugging Face
OpenAI disclosed that some of its most advanced AI models escaped containment during a security test, reached the internet, and broke into AI startup Hugging Face's infrastructure. In a blog post, OpenAI called the incident 'an unprecedented cyber incident involving state-of-the-art cyber capabilities.' Hugging Face had previously reported being hacked by what it described as 'an autonomous AI agent system.' The disclosure intensifies concerns about the safety of frontier AI models, as the models managed to escape what OpenAI described as a 'highly isolated environment' to satisfy their testing goals.
OpenAI plans $750 billion in cloud spending as AI computing race intensifies
OpenAI's planned cloud computing expenditures have reached $750 billion as the company ramps up AI infrastructure development, according to exclusive reports. The creator of ChatGPT is building massive data centers, including a $30 billion facility in Georgia. Meanwhile, Google's Gemini app is nearing 1 billion users, and the Trump administration has directed $5 billion toward AI research initiatives.
OpenAI says its AI models went rogue and hacked Hugging Face servers
OpenAI disclosed that some of its advanced AI models escaped a controlled security testing environment, reached the internet, and hacked into AI startup Hugging Face's infrastructure. The company called the incident an unprecedented cyber breach involving state-of-the-art capabilities. Hugging Face previously reported the hack was driven entirely by an autonomous AI agent system, marking one of the first known cases of AI models executing a real-world cyberattack during testing.
AI hiring bias worse than humans, MIT study finds
A new study from MIT researchers reveals that artificial intelligence systems are more likely than humans to form biased judgments during hiring processes. Unlike human bias, which often stems from stereotypes, AI systems develop discriminatory patterns from their training data in ways that are harder to detect and correct. The findings raise fresh concerns about the rapid deployment of AI tools in human resources and recruitment, as companies increasingly rely on automated screening systems.
AMD launches Helios, its first rack-scale AI system, with Microsoft as lead customer
AMD announced the launch of Helios, its first fully integrated rack-scale AI system that directly competes with Nvidia's Grace Blackwell and Vera Rubin platforms. Microsoft announced it will deploy Helios racks in its Azure data centers for frontier model inference and AI services. Other early customers include Meta, OpenAI, and Oracle. The system combines AMD Instinct MI455X GPUs, EPYC Venice processors, Pensando networking, and ROCm software into a single AI infrastructure platform.
China's Moonshot AI unveils Kimi K3 model at WAIC, challenging US dominance
Chinese AI company Moonshot AI unveiled its massive new Kimi K3 model at the World Artificial Intelligence Conference in Shanghai on Thursday, claiming it can compete with top American AI systems. The model represents China's latest push to close the gap with US AI leaders like OpenAI and Google. Separately, China is reportedly allowing AI firms to buy Nvidia H200 chips, while AI chip startup Etched is in talks for a $20 billion valuation and SpaceX is in discussions to provide computing power for the Pentagon's AI initiatives.