Last September, we spoke about what a possible OpenAI chip would mean for the industry. Now, that possibility has become reality after the company announced the launch of its very first AI chip.
In its blog post, OpenAI said, 鈥淥penAI and Broadcom today unveiled Jalape帽o, OpenAI鈥檚 first Intelligence Processor: an accelerator architected around OpenAI鈥檚 vision for the future of LLM inference, and the first AI accelerator in a multi-generation compute platform the companies are building together to make advanced AI faster, more reliable, and more accessible to more people.鈥�
The chip was designed from scratch by OpenAI, with its deep knowledge on LLMs, and
Broadcom and Celestica helped, as the post put it, 鈥渋ndustrialise the platform through chip implementation, board, rack system integration, high-performance networking and scalable production systems.鈥�
This now puts OpenAI in direct competition with chipmakers like NVIDIA.
What Is Jalape帽o, And What Does It Do?
The chip was designed for LLM inference workloads, which is the process where a trained AI model responds to prompts and performs tasks.
OpenAI said, 鈥淛alape帽o is designed with flexibility to work with all LLMs guided by OpenAI鈥檚 insights into the inference needs of current and future AI models across the industry. Engineering samples of the Jalape帽o chip are running ML workloads in the lab at production target frequency and power, including GPT鈥�5.3鈥慍odex鈥慡park.鈥�
Right now, Jalape帽o is in early testing, and deployment will begin by the end of the year.
This chip will help models like ChatGPT to run faster and better, instead of relying on external companies鈥� hardware to achieve this.
It鈥檚 also important to note that these chips aren鈥檛 the same as those that train AI. The ones that train process bigger volumes of data compared to these ones.
More from News
- Experts Comment: Is Britain’s Reliance On Foreign Technology Becoming A National Security Risk?
- Full Fibre Adoption Continues To Accelerate Across The UK
- OpenAI, Anthropic And Google Are Discreetly Building Their Own AI Standards Body 鈥� What Would That Actually Decide?
- How Have Digital Nomads Created A New Market For Businesses?
- England鈥檚 New Tourist Tax Has No National Cap 鈥� What Does That Mean For Hospitality Businesses?
- The EU Cyber Resilience Act Starts Today: Can Businesses Really Report A Cyberattack In 24 Hours?
- Can Binge-Watching Be Addictive By Design? Inside The State Lawsuit Against Netflix
- Could Australia鈥檚 Proposed Algorithm Rule Break The Current Echo Chamber That Is Social Media Today?
Where Is The Future Of AI Chipmaking Going?
This comes as NVIDIA chips are high in demand. OpenAI鈥檚 moves takes the costs and control into its hands directly, and more AI companies may follow.
Hock Tan, President and CEO, Broadcom, said, 鈥淥ur collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI.鈥�
鈥淭his is just the beginning of a multi-generation roadmap. By co-developing our industry-leading silicon directly with OpenAI, we are enabling the deployment of gigawatt scale data centers with Microsoft and other partners beginning in 2026.鈥�
鈥淭he world is moving to a compute-powered economy,鈥� said Greg Brockman, President and Co-Founder of OpenAI.
鈥淛alape帽o is part of our long-term full-stack infrastructure strategy to make compute more abundant, resulting in AI which is faster, more reliable, more affordable for people and businesses, and can be used to solve more important problems. By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access.鈥�
Basically, OpenAI Jalape帽o is a custom AI processor built to help OpenAI run its AI products faster, cheaper and at a bigger scale.
The company doesn鈥檛 only want to use it within the organisation – it also wants to sell Jalape帽o to external organisations who are looking for AI chips to speed up their processes.
It鈥檒l be interesting to see how this ends up impacting the semiconductor industry at large.
