Articles by tag: ai-hardware

NeoLogic quasi-CMOS: technological pioneer of WEF 2026

Israeli startup NeoLogic revolutionized chip architecture with quasi-CMOS. 3x reduction in power consumption. Find out who wins and who loses.

UK Sovereign AI: $1.5 Billion Plan and the Chip Battle

Analysis of the UK's plan to create sovereign AI hardware worth $1.5 billion. The role of AMD, startups Fractile, and the hidden battle for ARM. Read the details.

Nvidia and SK Hynix: multi-year agreement on HBM4 memory for AI

Analysis of the multi-year partnership between Nvidia and SK Hynix on the development of HBM4 and HBM4E. Who won, who lost, and why it changes AI infrastructure. Read the details.

POSTECH multifunctional transistor for AI devices: breakthrough or curiosity?

South Korean scientists created a transistor with 4x acceleration and 75% component reduction. Analysis of technology, prospects, and risks for Samsung, Intel, and TSMC.

NVIDIA Rubin AI architecture launched into production in the USA: analysis

Analysis of NVIDIA Rubin's quiet coup: why startups got chips first, who loses, thermal issues, and implications for the AI market. Read.

Nvidia and Microsoft ARM Computers: RTX Spark on Windows

Nvidia and Microsoft are preparing ARM PCs with the RTX Spark chip and AI performance of 1 petaflop. Learn about the implications for Intel, AMD, and the market. Read the details.

Meta is developing a wearable AI device: Zuckerberg's strategy

Meta has confirmed the development of a wearable AI device. We analyze the 'octopus' strategy, Hatch subscription, competition with Apple, and privacy risks. Read the details.

Apple M5 chip: 3nm process, 128 neural engine cores, AI revolution

Apple M5 with 3nm TSMC and 128 neural cores: AI 4x faster than M4. Why the chip is not needed by 95% of users, but Apple sells it to everyone. Detailed analysis.

HBM4: mass production by Samsung and SK Hynix for NVIDIA — problems and reality

Samsung and SK Hynix have started production of HBM4 for NVIDIA Rubin. Doubled bandwidth hides problems: warpage, SK Hynix lagging behind, and capacity reduction. Learn the truth about the AI memory race.

Quantum chip at room temperature: breakthrough in AI and photonic technologies

Prototype nanophotonic chip from Monash University operates at room temperature, processing quantum information and images. Learn about valleytronics technology and its advantages.

Polariton transistor: computing without electrons

Scientists have created a polariton transistor that operates on light. Learn how the 20 fJ/bit technology will change AI chips and data centers. Read the analysis!

AMD Helios vs Nvidia: war for AI racks

AMD launches Helios — an open AI rack against Nvidia. Find out why this is a data center market reshuffle, not just a new chip. Analytics and forecasts.

Nvidia and Corning: replacing copper with optics in AI infrastructure

How the $3.2 billion Nvidia and Corning deal is changing the fiber optics market for AI. Exclusive access, 10x capacity growth, and the end of copper cables. Read the analysis.

China's quantum computer with 1000+ qubits: breakthrough

Quantum computer on photonic chips with 3000+ qubits presented in Beijing. Learn how this will change the market and challenge IBM. Read analysis.

TPUv8 from Google: two chips for AI training and inference

Google splits TPUv8 into specialized chips for training and inference. Learn how this changes the architecture of AI infrastructure and competition with Nvidia.

Local AI on RTX 3090: mining build

Build on H510 Pro BTC+ and 3 RTX 3090 for gpt-oss-120b: 110 t/s inference, loading optimization. Tests, diagrams, costs for LLM developers.