OpenAI unveils Jalapeño chip to reduce dependence on Nvidia and make ChatGPT cheaper to run

OpenAI just made its biggest move against Nvidia -- and it could make ChatGPT cheaper to run

Meet Jalepeno, the chip that might be already running GPT-5.3-Codex-Spark OpenAI has unveiled Jalapeño, its first custom-built AI processor, developed with Broadcom specifically to run the large language models that power ChatGPT. Unlike the graphics processors (GPUs) that dominate today's AI infrastructure, Jalapeño was designed from the ground up for one job: answering user prompts as quickly and efficiently as possible. This move is huge for the future of ChatGPT as the AI assistant could become faster, cheaper to operate and more reliable over the next few years. Why OpenAI built its own chip Nvidia remains the undisputed leader in AI hardware, supplying the GPUs that train and run many of the world's most advanced AI systems. But those chips were designed to tackle a wide variety of computing workloads, not exclusively the task of serving billions of chatbot responses every day. Now Jalapeño takes a different approach. Instead of being a general-purpose processor, it's an inference chip, meaning its hardware is optimized specifically for generating answers after an AI model has already been trained. That's the part of the AI process you interact with every time you ask ChatGPT a question. According to OpenAI and Broadcom as reported by Reuters, the new chip was architected around the memory, networking and computing patterns used by modern large language models, allowing it to perform those workloads more efficiently than conventional hardware. Early internal testing suggests it delivers better performance per watt than today's leading AI accelerators, although independent benchmarks have not yet been released. Why this matters for ChatGPT users If OpenAI can answer more questions while using less electricity and fewer expensive chips, the economics of running ChatGPT begin to change. That doesn't necessarily mean your ChatGPT Plus subscription will suddenly become cheaper, but lower infrastructure costs could allow OpenAI to respond to prompts faster, support more users during peak demand, reduce outages caused by hardware shortages and roll out more capable AI models without dramatically increasing operating costs. In other words, beyond building a faster chip, OpenAI is making ChatGPT itself more scalable. OpenAI is becoming more vertically integrated Until now, OpenAI has largely relied on partners to provide the hardware powering its AI systems. By designing its own processors while continuing to build its own models and products, the company is gradually taking control of more of the entire AI stack. That's a strategy we've seen succeed elsewhere in technology. Companies that design both their hardware and software can optimize them to work together, improving performance while reducing long-term costs. OpenAI has described Jalapeño as the first step in a multi-generation compute platform, suggesting this won't be its last custom processor. To me, this looks more like an effort to reduce dependence on a single supplier while lowering one of OpenAI's biggest ongoing expenses. Nvidia still dominates AI training, and OpenAI continues to rely on Nvidia hardware across much of its infrastructure. Jalapeño is designed for inference, not to replace every GPU in OpenAI's data centers. The takeaway Interestingly enough, AI actually helped design the chip. The company says Jalapeño went from concept to production-ready design in just nine months, which is an unusually fast timeline for advanced semiconductor development. It's another reminder that AI isn't just writing code anymore, but increasingly, it's helping engineers build the hardware that future AI systems will run on. Although most users will never see Jalapeño, (they'll never need to), but they will . notice the results every time ChatGPT responds a little faster, serves a few more people or unlocks capabilities that were previously too expensive to run. Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok. Finally, you can visit our dedicated Tom's Guide Savings Squad hub for expert help on getting the best products for less.

OpenAI spices things up with Jalapeno chips: Why the Sam Altman-led AI startup is building its own processors

OpenAI said it designed the chip from scratch in nine months. Broadcom handled silicon implementation, networking, and connectivity, while Celestica developed the physical computing hardware of the chip, including its circuit board, hardware frame, and system integration. American artificial intelligence (AI) startup OpenAI and American chipmaker Broadcom on Wednesday unveiled Jalapeño, the startup's first custom AI chip built for inference, a process that enables AI models such as ChatGPT to serve users more effectively. What exactly is Jalapeño? The application-specific integrated circuit (ASIC), co-developed with Broadcom and system partner Celestica, is designed to deliver lower-cost, high-performance AI inference. An ASIC (Application-Specific Integrated Circuit) is a microchip customised for a specific function, rather than general purpose use. OpenAI said it designed the chip from scratch in nine months. Broadcom handled silicon implementation, networking, and connectivity, while Celestica developed the physical computing hardware of the chip, including its circuit board, hardware frame, and system integration. The company claims that the chip's engineering samples are already running machine learning workloads, including GPT-5.3-Codex-Spark, at target performance and power levels. The companies first disclosed the project last October, aiming at deployment of over 10 gigawatts of custom AI accelerators used for large language model (LLM) inference. Has OpenAI started using the new chip? No. Jalapeño marks the first generation of OpenAI's in-house compute platform, with initial deployment planned by the end of 2026. Why are such chips being built? Custom ASICs are alternatives to general purpose processors such as the graphics processing units (GPUs) produced by giants such as Nvidia. While Nvidia's H100 and A100 GPUs remain the industry's general-purpose AI processors, major cloud and AI companies are increasingly developing custom chips tailored to their own workloads. These custom chips claim to lower power consumption and reduce the cost of running AI models. As a result, tech companies such as OpenAI have turned to Broadcom and Marvell Technology, key providers of these custom silicon designs for data centres. Per a 2024 Reuters report, research firm 650 Group's Alan Weckel estimated that the market for custom data centre AI chips was expected to double to $20 billion in 2025. Separately, Needham analyst Charles Shi estimated the broader custom chip market was worth about $30 billion in 2023, or roughly 5% of global annual semiconductor sales. Does this mean OpenAI is moving away from Nvidia? OpenAI has been one of Nvidia's largest customers for AI chips, but surging demand for its models is pushing it to diversify its compute stack. Earlier this year, the company struck a deal with Amazon Web Services to use its Trainium AI chips and has also partnered with Advanced Micro Devices (AMD) and AI chipmaker Cerebras to expand its access to AI hardware. What does this mean for Nvidia? Nvidia is also exploring custom silicon. Per a report by The Information, the company was evaluating Intel's manufacturing technology for a processor that combines four GPUs into a single package, although it has not placed an order. The move builds on the Jensen Huang-led company's broader push into custom AI chips. Reuters reported in 2024 that Nvidia had created a dedicated unit to design bespoke AI processors for cloud providers and enterprise customers, seeking to tap the fast-growing custom chip market. What about OpenAI's rivals? Other AI companies are also moving beyond Nvidia. Per a CNBC report that came out in May, Microsoft is in talks to supply its custom Maia AI chips to the Dario Amodei-led AI startup Anthropic. Anthropic has also diversified its AI infrastructure, signing a 10-year, $100 billion deal with Amazon Web Services (AWS) to use Trainium chips and announcing support for Google's Tensor Processing Units (TPUs). Google, meanwhile, introduced its inference-optimised Ironwood TPU last year and, according to The Information, has ordered more than three million TPUs from Intel for delivery in 2028.

What OpenAI's Custom Jalapeño Chip Means for the Future of AI

OpenAI's new Jalapeño chip represents a significant step in the evolution of AI hardware, designed specifically to address the computational demands of large language models (LLMs). Developed in collaboration with Broadcom, this custom chip focuses on improving the efficiency and scalability of AI inference tasks, which are critical for generating outputs from input data. By reducing energy consumption and operational costs, Jalapeño aims to overcome the resource-intensive challenges of deploying LLMs. AI Grid explores how this hardware development aligns with OpenAI's broader strategy of reducing reliance on external suppliers like Nvidia, offering a glimpse into the company's vision for a fully integrated technology stack. Gain insight into how Jalapeño's architecture is tailored to optimize AI workflows, from its energy-efficient design to its potential for lowering costs by up to 50%. Discover the collaborative efforts behind its creation, including partnerships with Broadcom and TSMC and learn how OpenAI employed its own AI models to accelerate the chip's development. This explainer also provide more insights into the strategic implications of Jalapeño, examining its role in shaping OpenAI's gigawatt-scale infrastructure and the challenges that lie ahead in validating its performance claims. Purpose and Design of Jalapeño Jalapeño is specifically engineered to address the computational demands of AI inference, the process where AI models generate outputs based on input data. This stage is particularly resource-intensive for LLMs, which require vast amounts of processing power to deliver results efficiently. Jalapeño is designed to overcome these challenges by making inference faster, more cost-effective and energy-efficient. By aligning hardware with the specific requirements of its AI models, OpenAI aims to eliminate critical bottlenecks in AI deployment. This approach not only enhances performance but also reduces operational costs, allowing broader and more sustainable applications of AI technology. The Strategic Motivation Behind Jalapeño The development of Jalapeño stems from OpenAI's need to address the limitations of relying on external chip suppliers. Current hardware solutions, such as those provided by Nvidia, are often expensive and subject to supply chain constraints, which can hinder scalability. By creating its own chip, OpenAI gains greater control over its supply chain and the ability to tailor hardware specifically to its workflows. This move aligns with OpenAI's broader strategy of achieving independence across its technology stack. By reducing reliance on third-party suppliers, OpenAI ensures long-term sustainability, cost efficiency and the ability to innovate without external limitations. Here are more guides from our previous articles and guides related to OpenAI that you may find helpful. Performance and Efficiency: Key Differentiators Jalapeño is designed to deliver superior performance per watt, setting it apart from leading chips like Nvidia's Blackwell and Google's TPU. OpenAI has claimed that the chip achieves significant energy savings and faster processing speeds, which could lower costs and improve scalability for AI applications. Notably, the chip was developed in just nine months, a remarkably short timeline for such advanced hardware. While these claims are promising, they have yet to undergo independent verification. As such, the true extent of Jalapeño's capabilities remains to be seen and further scrutiny will be essential to validate its performance metrics. Collaborative Development and Industry Partnerships The creation of Jalapeño was made possible through collaboration with several industry leaders. OpenAI designed the chip's architecture, while Broadcom contributed its expertise in silicon engineering and networking. Manufacturing and assembly are being handled by TSMC and Celestica, making sure a streamlined production process. This collaborative approach highlights the importance of partnerships in achieving innovative advancements in AI hardware. By using the strengths of multiple organizations, OpenAI was able to accelerate the development of Jalapeño and ensure its readiness for deployment. AI-Assisted Design: A Innovative Approach In a notable innovation, OpenAI utilized its own AI models to assist in the design and optimization of Jalapeño. This approach demonstrates the potential of AI to streamline hardware development, reducing both costs and timelines. By using AI in this way, OpenAI not only improved the efficiency of the design process but also showcased the versatility of its technology in addressing complex engineering challenges. This method could serve as a blueprint for future hardware development across the industry, illustrating how AI can be used to tackle traditionally time-consuming and resource-intensive tasks. Strategic Implications for OpenAI Jalapeño represents more than just a technological advancement, it signifies a strategic shift for OpenAI. By controlling its hardware, OpenAI can potentially reduce operational costs by up to 50%, making its AI services more scalable and sustainable. This aligns with the organization's long-term vision of building gigawatt-scale AI infrastructure, capable of supporting the growing demand for AI applications. Additionally, the development of Jalapeño reinforces OpenAI's commitment to innovation and self-reliance. In an increasingly competitive industry, this move positions OpenAI as a leader in both AI software and hardware, setting the stage for future advancements. Future Plans for Jalapeño Jalapeño is the first in a planned series of custom chips. OpenAI intends to deploy initial servers powered by Jalapeño by late 2026, with full-scale implementation expected by 2029. These servers will form the foundation of OpenAI's gigawatt-scale AI infrastructure, developed in collaboration with partners like Microsoft. This infrastructure is designed to meet the growing demand for AI applications while making sure cost-effectiveness and sustainability. By investing in custom hardware, OpenAI is laying the groundwork for a future where AI technology is more accessible and efficient. Challenges and Uncertainties Despite its potential, Jalapeño faces several challenges. The high development costs associated with custom hardware underscore the need for cost-effective solutions to ensure long-term viability. Additionally, OpenAI's performance claims for Jalapeño have yet to be independently verified, raising questions about the chip's actual capabilities. Addressing these uncertainties will be critical for OpenAI to fully realize the benefits of its custom hardware. Independent testing and continued innovation will play a key role in determining the success of Jalapeño and its impact on the AI industry. A Fantastic Step for AI Hardware Jalapeño marks a significant milestone in the evolution of AI hardware. By addressing key challenges in performance, efficiency and scalability, the chip has the potential to reshape how LLMs are deployed and utilized. While hurdles remain, Jalapeño represents a bold step toward a future where AI infrastructure is more efficient, cost-effective and sustainable. As OpenAI continues to innovate, Jalapeño could set a new standard for custom AI hardware, influencing the trajectory of the industry for years to come. Media Credit: TheAIGRID Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.