OpenAI's GPT-5.5 Powers Codex on NVIDIA Methods, Remodeling Workflows

OpenAI has unveiled GPT-5.5, its newest AI mannequin, which powers the Codex utility and runs completely on NVIDIA’s GB200 NVL72 infrastructure. Over 10,000 NVIDIA staff, spanning roles from engineering to advertising and marketing, have already built-in GPT-5.5 into their workflows, reporting transformative productiveness features.

Codex, OpenAI’s agentic coding platform, leverages GPT-5.5 to streamline complicated duties. In response to NVIDIA, debugging cycles that used to take days now take mere hours, whereas function growth from natural-language prompts has develop into considerably quicker and extra dependable. The GB200 NVL72 techniques supporting this deployment ship 35x decrease price per million tokens and 50x greater token throughput per second per megawatt in comparison with previous-generation techniques, making enterprise-scale AI inference economically viable.

Jensen Huang, NVIDIA’s founder and CEO, highlighted the significance of this partnership in a company-wide e-mail, urging staff to embrace the Codex app. “Let’s bounce to lightspeed. Welcome to the age of AI,” he wrote.

Enterprise Safety at Scale

For enterprises contemplating AI adoption, safety is a important concern. NVIDIA’s deployment addresses this with sturdy safeguards, together with devoted digital machines for every consumer. These cloud-based sandboxes be certain that brokers function securely whereas sustaining full auditability. A zero-data retention coverage and read-only entry to manufacturing techniques additional defend delicate data.

This safe, scalable setup has made it simpler for NVIDIA to combine Codex throughout departments with out compromising knowledge integrity or operational workflows, setting a benchmark for enterprise AI adoption.

A Longstanding NVIDIA-OpenAI Partnership

The discharge of GPT-5.5 is the end result of a decade-long collaboration between OpenAI and NVIDIA. Their partnership started in 2016, when Huang personally delivered the primary NVIDIA DGX-1 AI supercomputer to OpenAI. The 2 corporations have since labored carefully on hardware-software co-design, with OpenAI offering suggestions that has formed NVIDIA’s product roadmap.

Notably, OpenAI is committing to deploy over 10 gigawatts of NVIDIA techniques to help its next-generation AI infrastructure. This formidable buildout will contain thousands and thousands of GPUs and positions NVIDIA as a foundational associate for OpenAI’s future developments in mannequin coaching and inference.

One tangible results of this collaboration is the primary GB200 NVL72 100,000-GPU cluster, which has already set new benchmarks for system reliability and scalability at frontier AI ranges. This infrastructure has been instrumental in delivering GPT-5.5’s capabilities.

Market Implications

Whereas OpenAI and GPT-5.5 are usually not publicly traded entities, the partnership with NVIDIA underscores the rising demand for superior AI infrastructure. NVIDIA’s dominant place within the AI {hardware} market, supported by partnerships like this, might additional improve its position as a important enabler of AI-driven innovation. Traders and trade stakeholders ought to monitor how these developments translate into broader enterprise adoption, because it might sign important development alternatives for NVIDIA.

For now, GPT-5.5’s efficiency inside NVIDIA highlights what’s attainable when cutting-edge AI fashions meet optimized {hardware}. As enterprises race to combine AI into their operations, this deployment affords a glimpse of what the way forward for work might seem like: quicker, smarter, and safer.

Picture supply: Shutterstock