Multiverse Computing has released Pulsar 16B, a 16.15 billion parameter AI model that delivers reasoning performance comparable to 30 billion parameter models but with half the parameters. Built on NVIDIA's Nemotron architecture and optimized through proprietary compression technology, Pulsar 16B offers improved speed and memory efficiency, making it suitable for deployment on lower-memory GPUs and single-node setups. The model excels in reasoning, coding, and tool-use benchmarks, and is available open-source on Hugging Face under the Apache 2.0 license, enabling enterprises to run advanced AI locally without cloud infrastructure costs.
Eco Wave Power, supported by NVIDIA AI infrastructure and digital twins, is developing technology to convert ocean wave energy into clean electricity using existing coastal structures. This approach allows wave energy to be generated near ports and i...
Super Micro Computer's stock surged 15.66% to $35.46 after unveiling a new Data Center Building Block Solutions blueprint based on Nvidia's Vera Rubin AI hardware. This development signals the company's resilience despite past regulatory and legal ch...
Nvidia introduced a warm-water cooling system that can almost eliminate water consumption inside data centers by using a closed-loop coolant system. This innovation significantly reduces on-site water use and improves efficiency by removing the need ...
Nvidia's stock has risen only 12% in 2026, underperforming the VanEck Semiconductor ETF's 84% gain, as investors shift attention to memory chips and AI infrastructure. The price to rent Nvidia's flagship B200 GPU, crucial for AI data centers, has fal...
Major tech companies are expected to spend about $750 billion on AI development this year, primarily funded through debt rather than cash reserves. This marks a shift from the past decade when these firms self-funded growth, making them less sensitiv...
