Today we're delighted to announce the release of two new models in the Granite Speech family: compact, 470M-parameter English speech recognition models that pair strong accuracy with unprecedented speed — over 12,600 RTFx on an NVIDIA H200 GPU, meaning...
Chips huggingface.co 2 sources signal 19 primary Hot
Overview Model Architecture Pre-Training SFT: Data Preparation & Quality Control Data Quality Control SFT Training Details Phase 2 SFT for the 30B Model Reinforcement Learning: A Multi-Stage, Multi-Environment Pipeline Training Methodology The Staged...
Chips huggingface.co 2 sources signal 18 primary
Why the usual healing methods fall short here Our approach Results QAH against QAT, head to head What this changes in practice Making a large language model smaller almost always comes with a cost.
Chips huggingface.co 2 sources signal 18 primary
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Chips huggingface.co 2 sources signal 17 primary
A look at the winning entries from the Gemma 4 Good Challenge.
Models blog.google 2 sources signal 17 primary
NVIDIA is bringing the next wave of RTX gaming to the Gamescom conference running this week in Cologne, Germany, with support for new games, anti-cheat technologies and increased visual quality.
Chips blogs.nvidia.com 3 sources signal 16 primary
Alabama's attorney general issued a subpoena to OpenAI on Monday as part of an investigation into how one of its AI agents escaped a supposedly secure testing environment and autonomously hacked another company last month.
Chips theverge.com 2 sources signal 16 verified
When an LLM engine process fails, the standard recovery path involves a cold restart. This requires loading weights into HBM from storage, compiling kernels,...
Models developer.nvidia.com 2 sources signal 15 primary
Nvidia worker indicted after Jensen Huang scolded Supermicro for AI server smuggling.
Startups arstechnica.com 2 sources signal 15 verified
No current X pulse in the 24-hour window.
No signals in the last 7 days.
Hardware developers and enterprise vendors are introducing custom inference silicon, liquid cooling, rack-scale systems, and high-throughput networking architectures to meet rising AI data center compute demands.
Capital and consolidation +7.6Investors and enterprise partners are deploying venture capital and forming regional commercial alliances across cybersecurity, data center energy efficiency, media generation, and vertical enterprise applications.
AI-linked equities are broadly positive, with Super Micro Computer +9.35%, Advanced Micro Devices +4.91%, NVIDIA +2.19% leading the tracked basket.
| Ticker | Company | Move |
|---|---|---|
| SMCI | Super Micro Computer | +9.35% |
| AMD | Advanced Micro Devices | +4.91% |
| NVDA | NVIDIA | +2.19% |
| META | Meta Platforms | +1.97% |