OpenAI presented its first internally designed AI chip, nicknamed "Jalapeño" and manufactured by Broadcom. The move marks its attempt to reduce its dependence on Nvidia, owner of the majority of the AI chip market. For Latin America, the relevant fact is what this can mean for the cost and access to artificial intelligence in the medium term. Here we explain what changes and why it matters.

The news, reported this week by TechCrunch and The Verge, is not a technical detail for experts: hardware is the bottleneck of the entire AI industry. Whoever controls the chips controls the cost and the speed of progress. That OpenAI takes this step reorders a board that until now was dominated by a single company.

What OpenAI announced

It's a processor designed by OpenAI specifically for its AI workloads, manufactured in partnership with Broadcom. The logic is the same one that led Google, Amazon, and Apple to design their own chips: optimizing performance for concrete tasks and no longer depending completely on an external provider.

Why the dependence on Nvidia matters

Nvidia dominates the market for chips to train and run AI models. That near-monopolistic position makes hardware more expensive and generates waiting lists. By designing its own chip, OpenAI seeks to:

• Reduce costs in infrastructure in the long term.

• Ensure capacity without depending on the availability of a third party.

• Optimize performance for its specific models.

To gauge the context of the AI hardware market, the coverage of specialized media like The Verge and TechCrunch has been closely following how the big tech companies compete for the chip supply chain.

How the AI chip market changes

More competition in the chip market usually translates, over time, into lower computing costs. And cheaper computing is what lowers the cost of access to AI for everyone.

What it means for LATAM

In the region, the cost of access to AI is a decisive factor for startups, SMEs, and professionals. If greater competition in hardware pressures prices downward, AI services could become more accessible, enabling more local use cases. The flip side: the manufacturing of cutting-edge chips remains concentrated in a few countries, so the impact won't be immediate.

To understand better how these underlying movements land in concrete tools, you can see how Microsoft boosts AI in Windows 11 with autonomous agents, another example of how the big tech companies integrate AI ever more deeply into their products.

Recommended Coderhouse courses

Understanding the economics of AI is key to getting ahead. To train, according to your level:

• To understand the landscape: the Introduction to Artificial Intelligence Course.

• For more technical profiles: the AI Engineering Course.

• To build products with AI: the AI Products Course.

Stay at the forefront of AI: train with Coderhouse and understand the trends that define the industry.

Frequently asked questions

What is OpenAI's "Jalapeño" chip?

It's the first AI processor internally designed by OpenAI, manufactured in partnership with Broadcom, designed to optimize its own workloads and reduce dependence on external providers like Nvidia.

Why does OpenAI want its own chip?

To reduce infrastructure costs in the long term, ensure computing capacity without depending on third parties, and optimize performance specifically for its AI models.

Is this going to make AI cheaper in Latin America?

Potentially yes, in the medium term. More competition in the chip market usually lowers the cost of computing, which can make AI services more accessible. The effect, however, is not immediate.

Does it mean Nvidia stops being relevant?

No. Nvidia continues to lead the market and be central to the industry. What changes is that the big players are starting to diversify with their own chips, reducing their exclusive dependence.