Graphics Processing Unit
9 reportsThe page treats performance per watt; memory and interconnects; and manufacturing constraints as distinct parts of Graphics Processing Unit. One line of support comes from power measurements, and a separate test comes from independent hardware analysis; together they clarify memory and interconnects, but the evidence is read with the caveat that vendor specifications may not represent sustained performance in real workloads.
Quandela Targets South Korea for Photonic Quantum Hardware
Quandela has identified South Korea as a potential Asian center for photonic quantum hardware, system integration and joint R&D, while emphasizing that the initiative remains a partner-search and development framework rather than a completed manufacturing agreement.
China's AI chip makers take on Nvidia's dominance in model training
Nvidia's lead in AI model training is under pressure as Chinese companies like Huawei, Cambricon, Moore Threads, and Biren Technology push to develop homegrown chips and software, aiming to reduce dependence on foreign hardware.
Figure and Nscale plan massive GPU buildout for robot AI training
Figure and Nscale have announced a multi-billion dollar partnership to deploy up to 100,000 NVIDIA GPUs for training general-purpose humanoid robot AI models, aiming to overcome the data and compute bottlenecks in physical intelligence research
Molex Backs Active Liquid Cooling Startup for AI Data Center Heat
Molex has invested in CAEPlus to develop BoundaryCool, an active liquid cooling system designed to manage the rising heat loads from advanced processors in AI and high-performance computing data centers
Nvidia Reduces Financial Guarantee for OpenAI's Ohio Data Center
Nvidia has scaled back its proposed financial guarantee for OpenAI's planned 10-gigawatt data center in Ohio, responding to investor concerns about risk exposure and the unprecedented scale of the project's financing requirements
QC Ware Tests Hybrid Quantum-Classical Chemistry on IBM Heron Processor
QC Ware has demonstrated a hybrid quantum-classical workflow for molecular energy calculations, integrating its Promethium platform with IBM's 156-qubit Heron superconducting quantum processor to analyze a complex metalloenzyme system
Meta's Muse Code Agent Tackles Extended Software Engineering Tasks
Meta has introduced Muse Code, a terminal-based coding agent powered by Muse Spark 1.2, designed to plan, write, and validate code across large repositories and sustain multi-step engineering work with reduced user intervention
SpaceX and Nvidia Plan Starmind AI1 Satellite for In-Orbit AI Processing
SpaceX and Nvidia have announced Starmind AI1, a satellite designed to run advanced AI workloads in low Earth orbit using Nvidia's latest space-grade processors, with ambitions for a vast orbital computing network
NVIDIA Donates DGX GB300 Supercomputer to Naval Postgraduate School
NVIDIA has provided its DGX GB300 AI supercomputer to the Naval Postgraduate School in California, aiming to support advanced research and education for military leaders in high-performance computing and artificial intelligence