AI Hardware
4 stories in this blend
The Neuron·3d agoCerebras Introduces CS-4 Server System for AI Inference
Hardware manufacturer Cerebras unveiled its CS-4 system designed specifically for high-speed model execution. The company claims the rack-scale architecture provides substantial speedups compared to traditional graphics processors.
The Neuron·5d agoNvidia signs $6 billion licensing agreement with AI startup Poolside
Nvidia has reached a non-exclusive deal worth $6 billion to license software from the AI coding company Poolside, while making a separate $1 billion investment. Under the arrangement, Nvidia will hire over 100 Poolside employees to support its internal projects, though the startup's leadership remains independent.
AI Daily Brief·Aug 15OpenAI releases ultra-high-speed mode for GPT-5.6 Sol model
OpenAI introduced an API setting called Ultrafast for GPT-5.6 Sol, achieving generation speeds of 750 tokens per second on hardware built by Cerebras. The speed enhancement focuses on applications where low response delay is essential, such as real-time audio and customer service.
AI Buzz!·Aug 14Running high-end artificial intelligence hardware can trigger millions in annual utility expenses
Data from market research firm Liftr Insights shows that operating a setup of 2,000 Nvidia graphics processing units carries immense ongoing electricity bills. In Dallas, running 1,000 H100 and 1,000 A100 chips costs about $2 million in power annually, compared to $2.1 million in Houston and $1.6 million in Austin. Despite high operational costs, major cloud providers continue to expand their use of these processors.
