Mac Apple Silicon: External AMD y NVIDIA GPUs Ahora Compatibles para IA (Vía TinyGPU)
Apple ha dado luz verde a una extensión de driver para GPU publicada por Tiny Corp, lo que facilita el uso de tarjetas gráficas dedicadas de AMD y NVIDIA con equipos Mac equipados con procesadores Apple Silicon (Arm). Es importante destacar que esta compatibilidad se da dentro del ecosistema de tinygrad y no representa un soporte nativo universal para macOS.
Esta aprobación por parte de Apple simplifica significativamente el proceso de instalación para desarrolladores. Anteriormente, era necesario desactivar la Protección de la Integridad del Sistema (SIP) para instalar manualmente los drivers de TinyGPU para GPU externas mediante USB4/Thunderbolt. Ahora, aunque el proceso aún requiere conocimientos técnicos, ya no es preciso deshabilitar esta protección clave del sistema.
La solución de TinyGPU permite el uso de GPUs AMD (RDNA3 o superior) y NVIDIA (Ampere o superior) en macOS 12.1 o posterior, requiriendo la instalación de TinyGPU.app y la activación de la extensión de driver, además de Docker Desktop para NVIDIA.
La principal finalidad de esta integración es el cómputo externo para inteligencia artificial y modelos de lenguaje grandes (LLM) a través de tinygrad/apps/llm.py. Esto significa que los usuarios no deben esperar mejoras en el rendimiento para juegos o creación de contenido general.
En resumen, esta noticia es de gran relevancia para desarrolladores de IA local, investigación y experimentación con modelos de IA en Mac Arm. Sin embargo, resulta de mucho menor interés para quienes busquen un mejor rendimiento gráfico general o revivir el concepto de eGPU para usos tradicionales.
Apple Silicon Macs: External AMD and NVIDIA GPUs Now Compatible for AI (Via TinyGPU)
Apple has approved a GPU driver extension published by Tiny Corp, facilitating the use of dedicated AMD and NVIDIA graphics cards with Mac computers featuring Apple Silicon (Arm) processors. It's crucial to note that this compatibility operates within the tinygrad ecosystem and does not signify native universal macOS support.
This approval from Apple significantly streamlines the installation process for developers. Previously, it was necessary to disable System Integrity Protection (SIP) to manually install TinyGPU drivers for external GPUs via USB4/Thunderbolt. Now, while the process still requires technical knowledge, it no longer demands disabling this key system protection.
TinyGPU's solution supports the use of AMD GPUs (RDNA3 or newer) and NVIDIA GPUs (Ampere or newer) on macOS 12.1 or higher, requiring the installation of TinyGPU.app and activation of the driver extension, in addition to Docker Desktop for NVIDIA.
The primary purpose of this integration is external computation for artificial intelligence and large language models (LLMs) via tinygrad/apps/llm.py. This means users should not expect performance improvements for gaming or general content creation.
In summary, this news is highly relevant for local AI developers, research, and experimentation with AI models on Arm Macs. However, it is of much less interest to those seeking improved general graphics performance or looking to revive the eGPU concept for traditional uses.
Fresh materials — Technology News

NVIDIA DLSS 5: Exclusive to GeForce RTX 50 Series GPUs and Beyond
NVIDIA utilized the SIGGRAPH 2026 event to provide a deeper explanation of how DLSS 5 will function. The significant news is that this technology is exclusively intended for GeForce RTX 50 graphics cards, built on the Blackwell architecture. Consequently, there has been no announcement of compatibi

NVIDIA GeForce RTX 3050 Found for Under $5 CAD at a Thrift Store – It Actually Works!
A Reddit user, known as Nervous, has made an almost unbelievable purchase: an MSI GeForce RTX 3050 VENTUS 2X 6GB graphics card for a mere 5 Canadian dollars. This incredible find, described as the "purchase of the year," was instantly bought from a second-hand store called Value Village. The de

CXMT Steps on the Gas: Its DDR5 Memory Now Reaches 8,800 MT/s with AMD Ryzen CPUs
Chinese manufacturer CXMT continues to make rapid progress in DDR5 memory technology, with its latest achievement being the 8,800 MT/s barrier on AMD platforms. The Chinese manufacturer Colorful has successfully pushed one of its iGame Shadow II kits to 8,800 MT/s on an AMD AM5 platform. This

NVIDIA brings RTX NTC technology to Windows 11 Arm ahead of RTX Spark chip launch
NVIDIA is gearing up for the release of its RTX Spark chips on the Windows 11 platform for ARM architecture. The latest advancement involves the release of the SDK for its RTX Neural Texture Compression (RTX NTC) technology, which now includes official compatibility with Windows ARM64. Previou

Microsoft Elimina Silenciosamente Recomendación de 32 GB de RAM para Windows 11: Casi Ningún PC se Vende con Más de 16 GB
Microsoft ha eliminado de su centro de aprendizaje varias páginas donde recomendaba 32 GB de memoria RAM para una experiencia fluida con Windows 11. Esta acción se ha realizado de manera sigilosa. Un documento eliminado, titulado 'Cómo optimizar tu PC Gaming' y publicado en noviembre de 2025,

Claude AI Bypasses HP BIOS RSA-2048 Checks, Unlocking 55 Hidden Settings for Laptop
While laptops offer portability, they often come with limitations compared to desktops, including restricted BIOS settings. However, a recent demonstration showcases how Artificial Intelligence can overcome these restrictions. By utilizing the AI model Claude, a user has successfully modified and