Apple’s rumored M7 Ultra targets 1.5TB of memory and Blackwell-class AI performance, report claims — monster 2028 offering would depend on memory shortage easing
 Apple’s rumored M7 Ultra targets 1.5TB of memory and Blackwell-class AI performance, report claims — monster 2028 offering would depend on memory shortage easing

Apple’s planned M7 Ultra chip is being designed to support up to 1.5 TB of unified memory and to push AI performance toward the class of Nvidia’s Blackwell accelerators.

, ,
Ireland’s data centers consumed nearly as much electricity as every home in the country combined in 2025 — server farms gulped 23% of national power despite years of grid restrictions

Ireland’s data centers consumed 23% of the country’s electricity in 2025, rising 10% in one year despite restrictions on new grid connections.

,
FCC approves orbital space mirrors, first test satellites will launch this year — large spacecraft reflects sunlight to Earth’s surface for construction sites, search-and-rescue lighting, and more

A startup that aims to bring sunlight on Earth after dark just received approval from the FCC to launch its experimental satellite. Critics say that the project could adversely affect astronomy and the environment, but the regulator says that their complaints are outside of its jursidiction.

Microsoft struggles to fulfill its 2030 sustainability promise amid carbon-heavy AI expansions — the company’s chief sustainability officer claims the target is still feasible
 Microsoft struggles to fulfill its 2030 sustainability promise amid carbon-heavy AI expansions — the company’s chief sustainability officer claims the target is still feasible

Microsoft’s carbon emissions jumped 25% in FY2025 as AI data center expansion outpaced sustainability gains, despite progress in water conservation and waste reduction.

,
Colibrì proof-of-concept gains frontier-level 1.5-TB AI model — novel approach runs on only 25GB of RAM and shows promise for local AI setups
 Colibrì proof-of-concept gains frontier-level 1.5-TB AI model — novel approach runs on only 25GB of RAM and shows promise for local AI setups

Colibrì proof-of-concept gets a frontier-level AI model running on only 25 GB of RAM and a modest CPU

,
Fake Go DNS scanner spread malware through over 200 GitHub repos — ‘Operation Muck and Load’ has published 700 malicious modules since January
 Fake Go DNS scanner spread malware through over 200 GitHub repos — ‘Operation Muck and Load’ has published 700 malicious modules since January

The module published its first version on January 24 this year and has since accumulated more than 1,200 versions.

,
Flock cameras mistakenly track car reviewer over ‘stolen’ tags — police ambush tester in store parking lot and detain him for an hour

Flock AI cameras failed to read the smaller digits on a non-standard New Jersey plate, leading cops to block in the driver on suspicion of driving a vehicle with “stolen” tags. It turns out the initial police report omitted the smaller numbers, resulting in several legitimate plates getting flagged by the system.

,
Apple sues OpenAI over alleged theft of trade secrets — claims company mentored incoming employees on bringing confidential information

Apple sued OpenAI, including its own former employees, over the theft of trade secrets as both companies build up AI hardware businesses.

,
SK hynix and TetraMem collaborate on experimental chip to bolster energy efficiency for edge AI devices — memristor-based in-memory SoC research leaves performance questions up in the air
 SK hynix and TetraMem collaborate on experimental chip to bolster energy efficiency for edge AI devices — memristor-based in-memory SoC research leaves performance questions up in the air

SK hynix, TetraMem, and the University of Southern California built a memristor-based in-memory computing system-on-chip for AI edge devices, achieving promising energy efficiency, but failed to demonstrate its full potential.

,
Anthropic says it can read Claude’s ‘thoughts,’ as detailed in new research paper — models observed to have a global workspace, revealing more of what makes LLMs tick

Anthropic has discovered an internal “J-space” for its Claude AI that displays similarities to human internal processing. While the AI developer anthropomorphizes it as thought, it may yet prove useful as a method of improving LLM honesty, oversight, and guardrails.

,