The upcoming A20 Pro chip boosts AI workload efficiency by adopting six-channel 96-bit LPDDR5X memory and new packaging, foregoing the latest LPDDR6 RAM standard.
Samsung begins mass production of its PM1763 PCIe 6.0 SSD, boasting unprecedented data speeds and power efficiency tailored for next-generation AI and HPC servers.
Chinese AI firm DeepSeek is reportedly developing proprietary inference chips to reduce reliance on Nvidia and Huawei, mirroring moves by U.S. AI companies seeking hardware control.
Microsoft has quietly transitioned thousands of user prompts in Excel and Outlook to its proprietary AI models, aiming to cut costs and reduce dependency on external providers.
Emerging agentic AI systems demand enhanced storage solutions, elevating storage technology from a back-end support role to a core element in AI infrastructure and efficiency.
NVIDIA revealed new details about its forthcoming Rosa CPU, featuring Rigel Arm v9.2 cores that boost single-thread performance and include a larger L2 cache within the same chip size.
Meta CEO Mark Zuckerberg admits AI development is lagging behind expectations amid staff cuts and reliance on competitors’ models, revealing internal turmoil.
NVIDIA rolls out a new Game Ready driver enhancing performance for DOOM: The Dark Ages Revelations and the fully remade Assassin’s Creed Black Flag Resynced.
NVIDIA refutes recent claims of delays for Rubin and Kyber AI chip platforms, affirming its product roadmap proceeds as planned without changes.
A new power supply listing reveals the NVIDIA GeForce RTX 5080 SUPER carries a 415W TDP, indicating higher power consumption alongside improved memory capacity and speed.
Sony Interactive Entertainment emphasizes AI’s role in boosting development speed and quality, using tools like synthetic voices and performance-capture enhancements.
Companies no longer rely on a single AI provider but blend multiple labs for specialized tasks, embracing flexible, cost-efficient AI strategies.