'It was awesome to see it work': OpenAI deploys GPT-5.5 Codex across Nvidia Blackwell systems – 50x efficiency boost and 35x cost reduction makes AI ‘viable at enterprise scale’ (www.techradar.com)

OpenAI has successfully deployed its GPT-5.5 Codex model across Nvidia's Blackwell systems, resulting in significant advancements for enterprise AI applications. This deployment has achieved a remarkable 50x increase in token output per megawatt and a 35x reduction in costs compared to the previous GPT-4o model. With over 10,000 Nvidia employees utilizing Codex, the AI is transforming workflow efficiency across various departments, from engineering to HR, allowing teams to complete complex coding tasks and generate end-to-end features from natural-language prompts within hours instead of days. This significant leap in performance underscores the growing viability of AI solutions at enterprise scale, as noted by Nvidia CEO Jensen Huang, who emphasizes that Codex can automate entire workflows rather than merely answering questions like traditional chatbots. As competition in the AI realm intensifies, particularly with Innovators such as Anthropic and its Claude Mythos model, the deployment of Codex highlights OpenAI’s commitment to pioneering applications that secure and enhance productivity in real-world business environments, thereby reshaping how organizations leverage artificial intelligence for everyday operations.