Cyberpunk Edgerunners 2 Netflix: premieres tomorrow with Hideo Kojima cameo
Categories: Anime
Cyberpunk Edgerunners 2 Netflix arrives tomorrow with Hideo Kojima reprising Hideyoshi Oshima from Cyberpunk 2077 in the new TRIGGER season.
Read moreOur Blog
blog_post.exe
19 Oct
blog_post.exe
19 Oct
DLSS mod for AMD Radeon: version 0.8.0 brings nearly the full NVIDIA DLSS suite
Categories: Tecnologia
Version 0.8.0 of the DLSS-NR on AMD mod adds Super Resolution, Ray Reconstruction, Frame Generation and Neural Rendering for Radeon cards.
Read moreblog_post.exe
19 Oct
Google Gemini Minecraft agents defeat the Ender Dragon in 27 hours with no human help
Categories: Tecnologia
DeepMind AI agents wrote 2,783 programs, died 605 times and coordinated on their own to complete Minecraft in 27 hours and 14 minutes.
Read moreblog_post.exe
19 Oct
US bans DeepSeek from civilian federal agencies: DHS orders removal and procurement ban
Categories: Tecnologia
The US Department of Homeland Security orders civilian agencies to remove DeepSeek products and third-party services based on its models from their systems and prohibits future acquisitions, citing risks of data transfer to China and security vulnerabilities.
Read moreblog_post.exe
19 Oct
Intel unveils GPU D-series with up to 480 GB memory for AI inference
Categories: Tecnologia
Intel presents the GPU D-series with up to 480 GB of LPDDR5X for AI inference, setting memory records in an air-cooled design.
Read moreblog_post.exe
19 Oct
NVIDIA launches Nemotron 3.5 Super VL and Bolt: open models for AI agents
Categories: Tecnologia
NVIDIA launches the 120B Nemotron 3.5 Super VL for enterprise agents and lightweight Bolt models for local RTX Spark PCs.
Read moreblog_post.exe
19 Oct
Viggle launches JST-2: physics-native world model generates 3D scenes and characters with motion in seconds
Categories: Tecnologia
Viggle presents JST-2, a physics-native world foundation model that creates expandable 3D scenes, characters with built-in motion dynamics, and animations from text or video, all in a unified architecture based on 3D atoms.
Read moreblog_post.exe
19 Oct
Inception's Mercury 3: Diffusion LLM hits over 1,000 tokens/s and outperforms GPT-6 Luna and Claude Haiku 5.5
Categories: Tecnologia
Inception Labs launches Mercury 3, its most capable diffusion-based LLM to date. The model offers a 1 million token context window, speeds above 1,000 tokens per second, and performance on par with or better than OpenAI and Anthropic speed-optimized models, with a 50% launch discount.
Read more