AMD’s controversial driver changes for older Radeon GPUs reveal a strategic shift in graphics card support. While day-one game patches continue, the move to maintenance mode signals deeper changes ahead for RDNA architecture owners.
Legacy UPS systems designed for predictable CPU workloads are failing under AI’s GPU demands. The power crisis is here, and retrofitting costs could be 2-3 times higher than planned upgrades. Here’s what’s really at stake.
AMD has clarified that it will continue supporting Windows 10 with driver updates despite Microsoft ending official support. This decision affects millions of gamers who haven’t upgraded to Windows 11. Here’s what this means for the gaming ecosystem and hardware longevity.
Microsoft’s staggering $34.9 billion capital expenditure in Q1 2026 reveals an aggressive infrastructure strategy. With $11.1 billion dedicated to data center leases alone, the company is racing to build what CEO Satya Nadella calls a “planet-scale cloud and AI factory.”
A 1,500-GPU liquid-cooled AI cluster is coming to Chicago in 2025, marking a significant expansion of AI infrastructure beyond traditional tech hubs. The deployment highlights growing demand for specialized computing power.
Pan-African technology group Cassava Technologies is accelerating its AI infrastructure expansion with strategic investments from Stanlib and Nvidia. The company plans to deploy thousands of GPUs across Africa by 2028, representing a major bet on the continent’s digital future.
HPE is pushing partners to embrace AI solutions with urgency, calling this a “do-or-die” moment for the channel. The company’s expanded HPE-Nvidia AI factory offerings aim to create new consulting and integration revenue streams.
A groundbreaking study reveals how intelligent client selection in federated learning systems can dramatically improve brain tumor detection accuracy. The approach achieves 82% dice coefficient for enhancing tumors while maintaining privacy across distributed medical data.
Computer scientists at UC Berkeley have demonstrated that AI systems can now outperform human experts in algorithm design. Their OpenEvolve AI coding agent developed a load balancing algorithm that runs five times faster than previous human-created versions, potentially transforming how systems research is conducted.
In what could signal a fundamental shift in how computer algorithms are designed, researchers at UC Berkeley have demonstrated that artificial intelligence systems can now outperform human experts in creating optimized algorithms. According to their recently published preprint paper, the team used an AI coding agent called OpenEvolve to develop a load balancing algorithm that significantly beats prior human designs.
NVIDIA is launching its powerful H100 GPU into orbit through a partnership with AI startup Starcloud. The collaboration aims to build the first space-based data centers, leveraging unlimited solar power and cosmic cooling. This ambitious project could solve Earth’s growing data center energy and cooling challenges.
NVIDIA is taking artificial intelligence capabilities beyond Earth’s atmosphere through a groundbreaking partnership with AI startup Starcloud, according to reports. The collaboration marks the first serious effort to establish data centers in space, with initial deployment scheduled aboard the Starcloud-1 satellite. Sources indicate this 60-kilogram platform will carry NVIDIA’s powerful H100 GPU, delivering what analysts suggest could be approximately 100 times the computing performance of current space-based systems.
