OpenAI and Broadcom unveil Jalapeño chip to power next-generation AI inference at scale

OpenAI and Broadcom announce chip designed for LLM inference at scale

OpenAI, the company behind ChatGPT and Codex and the models those tools utilize, and Broadcom, an established silicon supplier, have announced a new chip called Jalapeño, designed specifically for large language model inference in data centers. The chip is intended to be deployed at large data centers, both companies claim this is just the first generation in a long-term project that will see chips refined over time. Broadcom says that this ASIC (Application-Specific Integrated Circuit) was designed from scratch for LLM inference, based on "detailed insights" from the company's conversations with researchers at OpenAI, and that the chip's development was informed by OpenAI's own roadmap for future models and products. The design and production of the chip took nine months. The promise is that this chip is more specialized for the current needs of LLMs than those that inference systems currently run on in existing data centers. OpenAI claims that "early testing shows that Jalapeño will deliver performance per watt substantially better than current state-of-the-art," but notes that it is not done measuring performance, and that a "detailed technical report will be presented in the coming months." Until then, we don't have many details to go on. The company, which is known for its ChatGPT and Codex services and harnesses, hopes to ultimately own the full stack behind its models and products, reducing dependence on outside companies like Nvidia and ostensibly providing better performance or efficiency thanks to vertical integration. More generally, OpenAI and its competitors are interested in custom silicon because it's another way to potentially squeeze out more capacity amid a global compute crunch, as competing companies scramble for limited data center capacity. While Broadcom was already a successful chipmaker for customers building out compute infrastructure, it has seen substantial movement recently as it has built new business around providing custom chips to hyperscalers and the teams building frontier models during the current AI boom. Both companies claim Jalapeño chips will be deployed in data centers by the end of this year.

OpenAI unveils its first custom chip, built by Broadcom

On Wednesday, OpenAI unveiled its first custom-built inference processor, designed and manufactured in collaboration with Broadcom. Named Jalapeño, the new processor was designed specifically for the unique needs of OpenAI's inference systems. OpenAI's own AI models assisted in the development of the chip, the company said. While the chip is still being tested, OpenAI says early results show significantly better performance-per-watt than current state-of-the-art alternatives. The partnership was officially announced in October, but OpenAI's chip plans have long been rumored as a way to reduce the company's dependence on Nvidia's GPUs. Google and Amazon have both built custom chips to serve a similar purpose, often called "AI accelerators" -- silicon designed specifically to speed up machine learning workloads. OpenAI president Greg Brockman explained the company's approach to chip development on its in-house podcast, shortly after the Broadcom partnership was announced. "We have a deep understanding of the workload," Brockman said in the episode. "We've really been looking for specific workloads that are underserved, [and asking] how can we build something that will be able to accelerate what's possible?" Jalapeño is specifically designed for inference, the process of running pre-built AI models in response to user commands. In the announcement, OpenAI emphasized the chip's low operating cost when running real-time coding models. It's likely that more performance-intensive tasks like pre-training will still rely on Nvidia hardware, but even small reductions in inference costs could do a lot to improve the company's bottom line. Optimizing that inference system may prove to be a crucial factor in the economics of AI going forward -- and it's likely to take place at every level of the stack. OpenAI is already building agentic products like Codex and the models that power them, as well as data centers to run those models. Moving into purpose-built chips lets the company go even further in that process, as the company explained in its announcement. "OpenAI is not only developing frontier models or building products on top of them; it is designing the infrastructure underneath them: chip architecture, kernels, memory systems, networking, scheduling, deployment systems, and product experience," the company wrote. "Because OpenAI operates across the stack, each layer can be optimized around the same goal: making its models faster, more reliable, and more affordable for users."

OpenAI reveals its first AI processor: Jalapeño

OpenAI has just revealed a new "intelligence processor" chip for AI servers made in partnership with Broadcom. The chip, called Jalapeño, is designed to power current and future large language models, according to an announcement on Wednesday. Jalapeño is an ASIC (Application-Specific Integrated Circuit), meaning it's designed for a specific purpose: AI inference. With AI inference, models process a user's request to run an agent like Codex or offer a response from ChatGPT, while AI training involves a model consuming vast amounts of data to inform its responses. It comes just nine months after OpenAI revealed that it would team up with Broadcom to make its own chips, helping reduce the company's reliance on the GPUs offered by Nvidia that are in limited supply. In an interview with Reuters, Broadcom CEO Hock Tan says it matches the performance of Nvidia's Blackwell chips and Google's Tensor processing units. Microsoft, Meta, and Amazon are among the other AI companies that have also launched custom-designed AI chips recently to power their servers for either training or inference, while still trailing Nvidia's chips on overall performance. OpenAI calls Jalapeño the "first step in a multi-generation compute platform," which it expects to deploy by the end of 2026. "While OpenAI is still measuring final performance, early testing shows that Jalapeño will deliver performance per watt substantially better than current state-of-the-art," OpenAI says.

Broadcom and OpenAI unveil custom-built Jalapeño inference processor -- OpenAI's first chip is a massive reticle-sized ASIC built in an ultra-fast nine-month development cycle

OpenAI and Broadcom have introduced Jalapeño, a custom-built inference processor designed specifically for modern large language models and future agentic AI workloads, which is designed to deliver performance per watt they claim is higher than today's leading-edge hardware. OpenAI considers its hardware project a strategic one and envisions Jalapeño to be the first generation of its inference hardware. Not another AI accelerator OpenAI stresses that Jalapeño is a purpose-built inference ASIC and not a repurposed training accelerator or a general-purpose AI processor. OpenAI says the architecture of Jalapeño was designed based on its understanding of LLM behavior and is meant to address practical bottlenecks that matter for inference at scale, including costly data movement, balance between compute and memory resources, networking efficiency, and overall behavior. OpenAI also states that the design of the processor is meant to wed high throughput with low latency (which is why it uses a huge compute chiplet and HBM memory and not cheaper types of DRAM like many other inference accelerators), which will be particularly handy for reasoning and agentic workloads. In addition, OpenAI and Broadcom claim the processor is built to deliver higher effective utilization than conventional AI accelerators and deliver performance that is close to the theoretical maximum, which means very high efficiency both in terms of costs and in terms of power. Meanwhile, the companies did not disclose performance targets for their Jalapeño ASIC, so these claims should be taken with a grain of salt. Engineering samples are already operating in the lab at target clock speed and power (though Broadcom and OpenAI do not disclose details about this, either), and OpenAI says it is running machine learning workloads, such as GPT-5.3-Codex-Spark. The two companies also claim that early internal testing indicates that Jalapeño's performance-per-watt is substantially better than 'current state-of-the-art hardware,' although no hard numbers, benchmarks, memory configuration, or other details are disclosed, so again, we will have to take the claims with a grain of salt. In addition, one must bear in mind that while Jalapeño can purportedly beat existing AMD's Instinct MI350-series and Nvidia's Blackwell-based accelerators, it remains to be seen how competitive it will be against AMD's Instinct MI400-series and Nvidia's Rubin-based offerings. "Jalapeño was designed from the ground up for LLM inference using detailed insights from our close collaboration with OpenAI researchers," said Richard Ho, who leads OpenAI's hardware program. "We optimized the architecture around the kernels, memory movement, networking, and serving patterns that matter most for frontier AI models. Based on early testing, Jalapeño will efficiently execute our most important workloads close to the hardware's theoretical limits." A massive chip with six HBM modules While Broadcom and OpenAI did not disclose specifications of Jalapeño, they did show its wafer and packaging, so we can do a brief analysis. The package appears to contain one large compute chiplet surrounded by six HBM modules and another chiplet that likely packs input/output interfaces and is surrounded by two structural dummy dies. The wafer image does look like a Broadcom-style systolic-array-heavy accelerator, in the sense that it shows a very regular, repeated, columnar floorplan with what looks like replicated compute regions and fixed infrastructure macros. Yet, keep in mind that we are speculating, and the image is not clean enough to say that this is definitely Broadcom's standard TPU-like systolic array template with some perks from OpenAI, From the image alone, it is impossible to tell whether Jalapeño uses a true 2D systolic array, a set of 1D/2D matrix engines, a collection of vector or tensor tiles, or some other inference datapath. All we can say is that the die has a highly repetitive floorplan consistent with several kinds of tiled AI accelerator architectures. What we can tell from the image is the approximate die size of Jalapeño's compute chiplet based on the size of HBM3/4 packages (10.975 mm × 10.975 mm) that surround it. From what we can tell, the chiplet measures 25.46 mm (width) × 33 mm (height), which means that its die size is around 840 mm, which is very close to the reticle size of EUV lithography systems (858 mm). Given that the quality of the shot is poor, the die size we estimate cannot be 100% accurate, but we suspect it is close enough. The die size of Jalapeño's compute chiplet implies that it packs quite a lot of compute oomph, though, of course, we cannot make performance estimates based on this metric. Yet, it is safe to say that Jalapeño's compute die is considerably bigger than compute dies of other inference accelerators on the market and more resembles processors for AI training. Speaking of processors for AI training, we increasingly see multi-chiplet designs for these workloads as companies like AMD and Nvidia want to pack as much performance as possible. Meanwhile, the fact that OpenAI and Broadcom chose to go with a large compute chiplet possibly indicates that they wanted to reduce latencies by as much as possible. Designed in nine months The companies say the chip reached tape-out in just nine months and is slated for deployment beginning in late 2026, which represents an extremely fast turnaround time in ASIC design. It is unclear whether Broadcom and OpenAI extensively used artificial intelligence to define and then develop Jalapeño, though the companies admitted that they used OpenAI's models to speed up parts of the chip's design and optimization work. Typically, it takes 1.5 - 2 years to design an ASIC from scratch, so AI can shrink the development cycle. Another means to accelerate the design cycle is Broadcom's extensive reuse of its logic across different custom designs to deliver new chips faster than other companies. It is noteworthy that, according to the announcement, Jalapeño is designed to support not only OpenAI's own workloads but also present and future LLMs across the industry, which potentially lets OpenAI sell its hardware to third parties, assuming that it can get enough supply from Broadcom and TSMC. Meanwhile, the chief executive of Broadcom indicates that Jalapeño will be deployed at gigawatt-scale data centers with Microsoft and other partners starting this year, though it is unclear whether the processor will be used exclusively for OpenAI workloads or will be available for other tenants as well. "Our collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI," said Hock Tan, President and CEO, Broadcom. "This 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." Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

OpenAI gets chippy with Broadcom

OpenAI and Broadcom have teamed up - with a little help from some of the former's AI models - to develop the frontier model lab's very first inference chip, dubbed Jalapeño, the companies announced in a press release on Wednesday. Details of the spicily named silicon are scarce in the announcement, with the company admitting that it's running engineering samples of Jalapeño in its lab "at target frequency and power," but noting that it won't have any technical details to share until a report on its performance is released in the coming months. But that doesn't stop it from claiming that "early testing shows that Jalapeño will deliver performance per watt substantially better than current state-of-the-art." Alright then! OpenAI said that it designed Jalapeño itself, with Broadcom serving as its implementation and integration partner (i.e., they made the darn thing), but it wasn't humans alone who helped come up with Jalapeño's made-for-inference ASIC architecture: AI helped, natch, and the result is what OpenAI says is the fastest ASIC development cycle ever in the high-performance advanced semiconductor space. "Jalapeño was co-developed from initial design to manufacturing tape-out in just nine months," OpenAI claimed. "That speed reflects deep software-hardware co-development with OpenAI's engineering teams, Broadcom's silicon implementation expertise, and the use of OpenAI models to accelerate parts of the design and optimization process." In other words, AI is now helping design the chips it'll run on. Here's hoping they ironed out the hallucinations before heading to production. OpenAI explained in the announcement that Jalapeño is just the first of its AI accelerators, with the chip serving to define its "vision for the future of LLM inference," and one that will involve OpenAI controlling the entire stack behind its models and products. According to the release, OpenAI envisions a future where it doesn't just own the frontier models and the products built on top of them, but the infrastructure underneath as well. "Chip architecture, kernels, memory systems, networking, scheduling, deployment systems, and product experience" are all part of OpenAI's full-stack vision, which it said will enable it to make models "faster, more reliable, and more affordable." And more locked in, one would assume, like a proverbial walled garden. The Apple of the AI world, if you will. OpenAI is far from alone in developing its own silicon to help power AI - most of the giants in the space, including Amazon, Google, Meta, and Microsoft, have been building and using their own silicon for AI for several generations now, and OpenAI arch-rival Anthropic is reportedly considering a similar move. No telling, either, how OpenAI is intending to continue funding this capital-intensive initiative, given that it ran an operating loss of over $20 billion last year, according to leaked financials reported by Ed Zitron, and has apparently committed massive amounts ($600 billion? $1.4 trillion?) to infrastructure spending over the next few years. But hey - if we questioned AI economics, nothing would ever get built, would it? ®

Jalapeño is the first AI chip from OpenAI and Broadcom - Engadget

It's a spicy start for the AI company's move into chip production. OpenAI and Broadcom have unveiled the design for Jalapeño, their first jointly-made chip. The pair of companies announced plans to collaborate on a making a custom "AI accelerator" in October 2025. In its blog post today, OpenAI called Jalapeño its "first Intelligence Processor: an accelerator architected around OpenAI's vision for the future of LLM inference." In other words, the processor is designed to run its large language models. The AI company claims that so far, Jalapeño is offering "performance per watt substantially better than current state-of-the-art" in chip technology. The blog post did hedge that final performance testing is still underway, and a more thorough technical report on the chip's performance specs will be released in the coming months. The next stage will be to see Jalapeño put to use in data centers, with the initial deployment slated to begin in late 2026. This marks OpenAI's first foray into making the chips that it relies on for training and operating its ChatGPT models. By controlling more components across the process of running an LLM, the company says it can make its business more efficient and increase how much people will be able to effectively use ChatGPT tools. OpenAI also wrote that "Jalapeño was co-developed from initial design to manufacturing tape-out in just nine months," adding that the "speed reflects deep software-hardware co-development with OpenAI's engineering teams, Broadcom's silicon implementation expertise, and the use of OpenAI models to accelerate parts of the design and optimization process." Jalapeño is the first AI accelerator from the Broadcom and OpenAI partnership, and the two have said the collaboration is meant to be a "multi-generation compute platform" designed to "make advanced AI faster, more reliable, and more accessible to more people."

OpenAI unveils first chip as part of Broadcom deal in effort to 'build the full stack'

OpenAI and Broadcom on Wednesday unveiled their debut custom chip, called Jalapeño, marking the ChatGPT maker's first entry into artificial intelligence silicon. The chips will be made by Broadcom and used by OpenAI for inference, the compute-intensive process of serving its AI models to users in ChatGPT and other applications. It's a major step in OpenAI's plan to "build the full stack behind its models and products," according to the press release. "By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access," Greg Brockman, OpenAI's president, said in the statement. Since OpenAI kick-started the generative AI boom in 2022, the company has been one of the biggest buyers of Nvidia's pricey graphics processing units, the key piece of infrastructure for building AI models and running large workloads. But OpenAI is experiencing such an explosion in demand that it needs other sources of advanced silicon. Earlier this year, OpenAI forged a deal with Amazon Web Services that includes use of the company's Trainium AI chips. OpenAI has also signed agreements with Nvidia rival Advanced Micro Devices and with AI chipmaker Cerebras, which held its initial public offering in May. In October, after 18 months spent working together, OpenAI and Broadcom went public with plans to develop and deploy racks of OpenAI-designed chips starting late this year, ultimately aiming to build enough to require 10 gigawatts of power.

OpenAI Distances Itself From Nvidia With Jalapeño, Its First In-House AI Chip

OpenAI fired the starting gun for the AI race in late 2022 with the launch of ChatGPT and has since become one of the most valuable startups in the world, reaching hundreds of millions of users every day. Its success, however, has depended on a web of partnerships with other companies that make the silicon chips upon which its AI systems depend. On Wednesday, OpenAI took its biggest step yet towards full technological independence with the unveiling of a new AI chip, dubbed Jalapeño, that it developed in collaboration with U.S. chipmaker Broadcom. In a blog post, the company described Jalapeño as an "Intelligence Processor" and said it moved OpenAI closer to becoming a "full-stack" AI platform, which is industry-speak for controlling the production of all of the hardware and software components necessary to build models and make them accessible to users. Jalapeño is not a GPU -- the class of chip made famous by Nvidia and the one most commonly used by tech developers to train and run AI models -- but an application-specific integrated circuit, or ASIC. As its name suggests, an ASIC is designed to perform more specialized tasks than a GPU, which is more general-purpose. The post also said that by being able to tap into its own supply of in-house chips, OpenAI would be able to deliver cheaper products and minimize wait times when demand is particularly high. "By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access," OpenAI president and cofounder Greg Brockman said in a statement. The company currently buys silicon mainly from Nvidia but also has partnerships with Amazon, Advanced Micro Devices (AMD), and Cerebras. In October, OpenAI and Broadcom announced their plan to build enough racks of custom-made AI chips to supply ten gigawatts of power, enough to power roughly seven and a half million homes. The companies now plan to implement Jalapeño chips into a full-scale production setting: "This is just the beginning of a multi-generation roadmap," Broadcom CEO Hock Tan said in a statement. "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." In its announcement on Wednesday, OpenAI said the development of Jalapeño took nine months from start to finish, marking "what we believe to be the fastest ASIC development cycle ever achieved in high-performance advanced semiconductors." That speedy development process was enabled in part by OpenAI's own AI systems, according to the blog post. "The same models served to users are helping improve the infrastructure used to run future models," the company wrote. "If AI can help engineers design better chips faster, it can lower the cost of compute across the industry and help democratize access to advanced AI." AI is also increasingly being used to help software engineers train new and more powerful models, a process known in tech circles as "recursive self-improvement." The operative word there is "recursive": If AI is able to continually improve its own abilities by writing and altering its underlying code, that could in theory allow it to become so sophisticated that it escapes human control -- an "intelligence explosion" with potentially catastrophic consequences for humanity. Both OpenAI and Anthropic have called for the creation of an international oversight committee to prevent such an outcome, if necessary, by enforcing an industry-wide slowdown on the development of new models.

OpenAI's Jalapeño chip: a way out from Nvidia

OpenAI just showed off Jalapeño, its first home-grown AI chip, built with Broadcom. It is designed for inference, not training. And it is the clearest sign yet that the company most dependent on Nvidia wants a way out. OpenAI has a chip now. On Wednesday it unveiled Jalapeño, the first piece of silicon it has designed itself. It is a pointed answer to a question that has hung over the company for years. What happens when the world's biggest buyer of AI compute decides it no longer wants to rent all of it from Nvidia? The chip was built with Broadcom, Axios first reported. OpenAI did the core design. Broadcom brought the connectivity and networking know-how, plus its Tomahawk switching silicon. A third partner, Celestica, handled the boards and racks. OpenAI is already running the first samples in its labs. There they answer Codex queries and run workloads for a model it calls GPT-5.3-Codex-Spark. OpenAI is not trying to replace Nvidia overnight. It is trying to stop being a captive customer. That is the real story here. What Jalapeño actually is Jalapeño is built for inference: the day-to-day work of answering user queries. It is not built for training new models. OpenAI calls it an "Intelligence Processor". It stresses that this is a blank-slate design, not a general-purpose accelerator bent to fit the job. The pitch is efficiency. Early testing, the company says, shows performance per watt "substantially better than current state-of-the-art". Thermal behaviour also came in better than expected. Those are OpenAI's own numbers, and a full technical report is still months away. For now, the claim to watch is narrow but real. A chip tuned for one job can beat a flexible one at that job. Inference is where AI meets actual users. So even small gains in cost and speed compound fast across hundreds of millions of daily queries. The plan is to put Jalapeño to work later this year. Broadcom expects the first chips in commercial use at Microsoft and other partners by the end of 2026, though OpenAI says real volume arrives next year. The longer goal is bigger still. OpenAI wants its custom chips powering 10 gigawatts of compute by 2029, roughly the output of ten nuclear reactors. A nine-month sprint, designed partly by AI The timeline is the flex. OpenAI and Broadcom say they took Jalapeño from first design to manufacturing tape-out in nine months, which they believe is the fastest such cycle ever for an advanced, high-performance chip. Tape-outs at this level usually take far longer. Part of how they did it carries a neat twist. OpenAI used its own models to speed up parts of the chip design. The same systems people query through ChatGPT helped build the hardware that will soon run them. If AI can genuinely help engineers design better chips faster, that lowers the cost of compute for everyone, which is exactly the kind of self-reinforcing loop OpenAI likes to talk about. It also helps explain the recent rush of startups using AI to design chips. Why build your own chip at all The motive is control as much as cost. "This gives OpenAI full stack control," said Richard Ho, who leads the company's hardware programme. OpenAI now designs the model, the software, the serving systems and, increasingly, the chip underneath. Owning each layer lets it tune the whole thing toward one goal: cheaper, faster intelligence. Broadcom chief executive Hock Tan put the case more bluntly. "At the end of the day, you cannot, should not rely on some other third-party GPU to do it for you, because it's such a key part," he said. The not-so-subtle target is Nvidia, whose chips have powered almost all of OpenAI's training and inference to date, and whose ecosystem lock-in is precisely what large customers now want to loosen. OpenAI joins a crowded club OpenAI is late to a party its biggest rivals threw years ago. Google has its TPUs, Amazon its Trainium and Graviton lines, and Microsoft its Maia accelerators. Each pairs custom silicon with Nvidia chips rather than replacing them outright. Anthropic is exploring its own chips too. The logic is shared across all of them: at this scale, designing your own silicon is cheaper than paying Nvidia's margins forever. One name keeps recurring on the other side of these deals. Broadcom now sits behind a striking share of the industry's custom accelerators, from Google's to Jalapeño, and recently struck a huge compute pact with Anthropic and Google. The company has quietly become the kingmaker of the post-Nvidia chip scramble, supplying the connectivity and manufacturing muscle that the AI labs lack. OpenAI has already started spreading its bets. Beyond Nvidia, it recently began using Cerebras chips for inference, part of a wider challenge to Nvidia in inference specifically, where rivals see the best chance to break the grip. Jalapeño turns that diversification into something OpenAI owns outright. The case for caution A first chip is not a finished strategy. Jalapeño handles inference, not training, and training is where Nvidia's lead is hardest to challenge. OpenAI concedes Nvidia remains a key partner there. So this is diversification at the edges, not a divorce. The performance claims also rest on OpenAI's own early testing, with the detailed report yet to come. Vendor benchmarks at launch deserve a raised eyebrow until independent numbers land. Building chips is slow, capital-hungry and unforgiving work, and a nine-month tape-out is a long way from reliable production at gigawatt scale. None of that makes the move wrong. It makes it a first step. The open question is whether OpenAI can really build its way out from under Nvidia, or whether owning a slice of its own silicon simply gives it leverage at the negotiating table. Either way, the company that buys more AI compute than almost anyone has decided it would rather make some of it. The rest of the industry will be watching how well Jalapeño holds up under real load.

OpenAI's new 'Jalapeno' chip is the company's first step towards the future

It's called "Jalapeno" and the company says it offers better performance-per-watt than current state-of-the-art chips. OpenAI has today unveiled its new chip developed in collaboration with Broadcom. The chip is called "Jalapeno," and it's the company's first "Intelligence Processor." It's a key step in OpenAI's plans to build the full stack for its services. The ChatGPT parent says that the chip is built around its vision for the future of LLMs and is the first step in the "multi-generation compute platform" the two companies are building together. The chip was designed by OpenAI, and as such, it takes into account the company's future plans for its products and models. Jalapeno is already running machine learning tasks at OpenAI's labs. The chip is meant to perform LLM inference, and OpenAI's aim is to create something that can offer performance similar to leading AI accelerators while reducing latency. In fact, the company's early testing suggests the new chip offers better performance per watt than current state-of-the-art options. Building the infrastructure that powers its products can prove hugely advantageous for OpenAI. With such deep control, the company can optimize its models, such as the upcoming GPT-5.6, to be faster and more affordable. It can also help improve training efficiency and allow OpenAI to develop and power more advanced AI models. Plus, it can reduce the company's dependence on chips sourced from third parties such as NVIDIA. The entire process, starting from the initial design to the manufacturing tape-out reportedly only took nine months, and the company used its own AI models to handle some parts of the process. OpenAI says that it's planning to deploy Jalapeno by the end of 2026, and will continue developing more such chips for its AI needs in the coming years. We'll get to know more details about the new processor once OpenAI releases its technical report in the coming months. If these initial claims prove accurate, it could mean we'll get to see much faster AI development from OpenAI.

OpenAI unveils Jalapeño chip for large-scale inference workloads

OpenAI has unveiled its first custom AI accelerator, called Jalapeño, marking the company's move into chip design as it looks to reduce the cost and improve the efficiency of running large language models (LLMs). Developed in partnership with Broadcom and Celestica, the chip is designed specifically for AI inference -- the process of generating responses from trained AI models. OpenAI said early testing shows Jalapeño delivers significantly better performance per watt than current state-of-the-art AI accelerators, though detailed benchmarks will be released later. The announcement expands OpenAI's efforts to control more of the infrastructure behind its products. In addition to building models and applications such as ChatGPT and Codex, the company is now designing the hardware that powers them. Engineering samples of Jalapeño are already running machine learning workloads in the lab, including GPT-5.3-Codex-Spark, at production target frequency and power levels, according to the company. Unlike general-purpose AI accelerators adapted for multiple workloads, Jalapeño was built specifically for LLM inference. OpenAI said the architecture was designed around the compute, memory, networking, and serving requirements of modern AI models. The company claims the chip reduces data movement while balancing compute, memory, and networking resources to improve hardware utilization. Broadcom contributed silicon implementation and networking technologies, including its Tomahawk networking platform. "Jalapeño 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," said Greg Brockman, President and Co-Founder of OpenAI. Richard Ho, who leads OpenAI's hardware program, said the accelerator was optimized around the workloads most important for frontier AI systems. "Based on early testing, Jalapeño will efficiently execute our most important workloads close to the hardware's theoretical limits," Ho said. The chip is also intended to support future LLMs across the broader AI industry, not just OpenAI's own models. According to the companies, Jalapeño was developed from initial design to manufacturing tape-out in just nine months. OpenAI described the effort as potentially the fastest ASIC development cycle achieved for a high-performance advanced semiconductor. The development process involved extensive software-hardware co-design between OpenAI and Broadcom engineers. OpenAI also said its own AI models were used to accelerate portions of the chip design and optimization workflow. "Our collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI," said Hock Tan, President and CEO of Broadcom. The companies plan to deploy the accelerator at gigawatt-scale data centers beginning in 2026. Jalapeño is the first product in what OpenAI describes as a multi-generation compute platform that will combine OpenAI-designed accelerators with Broadcom networking and connectivity technologies and Celestica's system integration expertise. OpenAI said improvements in inference efficiency could translate into faster ChatGPT responses, lower AI operating costs, and more reliable access to advanced AI services as demand continues to grow.

OpenAI has revealed its first custom chip, and Nvidia should be worried

OpenAI has revealed its first custom chip after announcing plans nine months ago, and the Broadcom-built design promises to upturn the AI industry -- particularly for Nvidia, one of the company's most important hardware suppliers. The newly introduced Jalapeño is billed as an Intelligence Processor, and is meant to handle AI inference (that is, applying machine learning to new tasks) with a performance-per-watt ratio that is "substantially better" than the industry's best. The silicon is optimized for frontier AI large language models (LLMs) like GPT, and OpenAI claims it's a true "blank-slate" chip designed for this purpose rather than a rehashed accelerator. That's a not-so-subtle dig at companies like Google, which originally made its Tensor chips for pre-LLM AI work. The chip is currently at the manufacturing tape-out stage, where the completed design is sent out for initial manufacturing. OpenAI and Broadcom claim the nine-month development is the "fastest" yet for an app-specific integrated circuit (ASIC), and that they're on track to put Jalapeño into "gigawatt scale" data centers from Microsoft and others starting later in 2026. The partners add that this is the start of a "multi-generation roadmap," suggesting that performance and efficiency should improve. How will OpenAI's custom chip transform AI? Cheaper and faster at Nvidia's expense OpenAI claims Jalapeño will lead to improvements in cost, performance, and reliability that you'll notice every day. You could get quicker ChatGPT answers, run more advanced Codex computer tasks, and get more consistent access during high-demand periods. Prices might also stay in check. However, the custom chip could also address some mounting problems in the AI field. Current LLM technology consumes massive amounts of power, to the point where governments are either requiring independent electricity for new data centers or even banning those server farms. Jalapeño can theoretically reduce that power draw and limit future demand increases. OpenAI is close to turning ChatGPT into a super app -- here's what it will do You could create and code in one place. Posts 2 By Jon Fingas There's no guarantee this will play out as promised. Jalapeño might improve efficiency, but OpenAI could still end up consuming more power if it scales aggressively or rolls out more computationally-intensive models. Competitors and hardware partners will have reason to be worried, at least. OpenAI will finally have custom AI accelerators like Amazon and Google. Moreover, OpenAI has heavily relied on Nvidia GPUs to speed up AI processing. Jalapeño might not completely replace Nvidia hardware, but the brand could have to scramble for new customers and features to make up for potential losses. ChatGPT+ What's included? Unlimited conversations, faster response speed, priority access, and more Brand ChatGPT Try for Free Expand Collapse

OpenAI unveils first custom AI inference chip, Jalapeño, with Broadcom -- and its development was sped-up with OpenAI's own models

OpenAI and Broadcom this morning unveiled their first custom AI accelerator chip named "Jalapeño," positioning it is as a purpose-built processor for large language model (LLM) inference, rather than the more general GPUs offered by the likes of Nvidia or AMD. According to its creators, Jalapeño is designed to support workloads behind ChatGPT, Codex, the API and future agentic products, though notably, Broadcom's news release positions it as a product that could be available to external AI firms as well -- "built from the ground up for current and future LLMs across the industry." [Emphasis mine.] Jalapeño's engineering timeline set a blistering pace for the semiconductor industry, moving from early schematics to fabrication readiness within a brief nine-month window, when new processor development cycles are typically measured in years. The companies attributed this speed to a deep software-hardware co-development process that actively used OpenAI's own models to accelerate parts of the chip design. After receiving an early physical model on Wednesday, OpenAI outlined plans to begin rolling out these processors across active data centers by the end of this year. OpenAI says it has already begun testing running at least one of its prior generation models, GPT‑5.3‑Codex‑Spark, on the chips at a production workload, though in a test environment. This release marked a major strategic shift for the ChatGPT creator as it attempted to build the full computational stack required to make advanced AI faster, more reliable, and more accessible. There remain, of course, many outstanding questions -- including how the new Jalapeño chip performs compared to direct competitors, its costs, and its manufacturing viability. Why OpenAI Built an ASIC To understand why OpenAI is moving into chip design, it helps to look at the architecture. Jalapeño is an Application-Specific Integrated Circuit, or ASIC. Unlike a GPU, which can handle many types of workloads, an ASIC is tuned for narrower uses, as industry experts note. That narrower focus can make it cheaper and more efficient for specific AI tasks, though less adaptable than Nvidia-style GPUs. In Jalapeño's case, OpenAI is starting from a clean design focused on modern LLM serving, instead of adapting a broader accelerator to fit its needs. The company says the architecture is shaped by its experience running large-scale AI products and is meant to reduce unnecessary data movement while better matching compute, memory and networking resources. Broadcom is contributing core silicon implementation and networking technology, including Tomahawk networking silicon, while Celestica is helping with board, rack and system integration. The goal is to move the chip closer to its practical performance ceiling in real workloads, not just improve theoretical benchmarks. However, OpenAI's pivot into proprietary hardware is not just as a quest for technical supremacy: it may also make its core unit economics far more sustainable. Audited financial documents posted recently by AI critic and AI public relations specialist Ed Zitron revealed that while OpenaAI generated an impressive $13.07 billion in revenue throughout 2025, its total operational expenses for the year ballooned to $34 billion, resulting in an operating loss of nearly $20.92 billion. The primary culprit behind this cash hemorrhage involved pure compute requirements, though more is likely due to training than inference. In 2025 alone, research and development costs -- driven largely by the infrastructure required to train and serve massive language models -- accounted for $19.18 billion, or approximately 56 percent of the company's entire spending footprint. Furthermore, OpenAI reportedly paid Microsoft over $10.59 billion just for R&D and compute infrastructure last year. Still, as OpenAI lays the groundwork for a heavily anticipated public offering in 2026, the Jalapeño inference chip may offer some reassurance to private investors and public markets that OpenAI has a plan for digging itself out of the financial hole and moving toward profitability. If it can drive down the costs of AI inference, then maybe it can recoup some of the losses spent on costly training runs. "By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access," said Greg Brockman, OpenAI's president and co-founder, in a statement included in Broadcom's release. What Does This Mean for Nvidia and All of OpenAI's Other Chip Providers? The introduction of Jalapeño immediately raises questions about OpenAI's strategic positioning within the fiercely competitive semiconductor and GPU market. Since kicking off the generative AI boom in late 2022, OpenAI has remained one of the largest customers of GPU market leader Nvidia's premium products, but has also taken billions in investment dollars from the firm (engendering accusations of "circular dealing"), and expanded to work with other rival chipmakers to fuel its appetites. * Nvidia: In February 2026, Nvidia finalized a $30 billion direct investment into OpenAI as part of a massive $110 billion funding round.This deal secured an agreement to deploy 10 gigawatts of computing systems -- including 3 gigawatts of dedicated inference capacity and 2 gigawatts of training capacity -- utilizing Nvidia's next-generation Vera Rubin platform. * Amazon Web Services (AWS): As part of the same February 2026 funding round, Amazon invested $50 billion into OpenAI. This deal included a commitment for OpenAI to consume approximately two gigawatts of AWS's proprietary Trainium computing capacity over the next eight years. * Advanced Micro Devices (AMD): OpenAI signed agreements with Nvidia's chief hardware rival, AMD for the former's usage of the latter's AMD Instinct™ MI450 Series GPUs. * Cerebras: The company also struck a pact with Cerebras, an AI chipmaker that executed its initial public offering in May 2026. The Gigawatt Future This sprawling web of vendor agreements highlights the sheer scale of OpenAI's infrastructural ambitions. The ultimate goal of the OpenAI and Broadcom partnership involves deploying gigawatt-scale data centers with Microsoft and other partners beginning in 2026 -- that is, data centers with compute requiring energy on the order of cities. For Broadcom, the partnership acts as a massive reputational catalyst. The company has been among the biggest beneficiaries of the generative AI boom, helping hyperscalers and frontier labs engineer custom silicon. Broadcom shares reflect this momentum, demonstrating an 18% year-over-year increase in the first part of 2026 and a nearly 7X boost since the end of 2022, according to CNBC. Ultimately, Jalapeño confirms that OpenAI believes it is ready to move beyond software and code into the realm of real-world, custom hardware. By controlling the physics of its inference pipeline -- while simultaneously leveraging the capital and hardware of Nvidia, Amazon, AMD, and Cerebras -- OpenAI is attempting to rapidly rewrite its future unit economics of AI.

Broadcom and OpenAI debut Jalapeño Intelligence Processor, plot an Apple-like move to 'build the full stack'

The AI arms race just got a little bit spicier as Broadcom and OpenAI target Nvidia * Broadcom and OpenAI reveal custom AI chip called Jalapeño * It is the first of a family of processors designed from ground up for inference and agentic AI * Global rollout is expected in 2027 Broadcom and OpenAI have finally announced their first custom chip, dubbed Jalapeño, designed from ground up by OpenAI for ChatGPT, Codex and 'future agentic products'. JIP, for Jalapeño Intelligence Processor, is inference-native and little is known about the hardware. A closer look at the packaged chip (lower left of the picture below) leads me to believe that the ASIC silicon is surrounded by eight HBM stacks to reduce latency to a minimum. OpenAI disclosed that it only took nine months for the chips to be designed and delivered, something that usually takes years, especially with an entirely new architecture. Company President Greg Brockman highlighted the role of ChatGPT as a virtual chip architect, which 'accelerated the tape-out', which he described as surprising. Jalapeño Intelligence Processor The 300mm wafer that both CEOs are holding will generate about 50 to 60 ASICs. Engineering samples of Jalapeño have been running ML workloads including GPT-5.3-Codex-Spark, OpenAI's first model designed for real-time coding and designed to be used with Cerebras's huge WSE (Wafer Scale Engine). OpenAI is following the footsteps of Apple, AWS, Google and many others that want to control the hardware, the software and the infrastructure - the so-called "full stack". Elon Musk's SpaceX wants to go a step beyond by being the only one that also builds the actual chips. While the focus is on Jalapeño, OpenAI will also partner with Celestica and Broadcom to build the rack systems and the network infrastructure that surrounds the intelligence processor. This should lead to much lower prices, more feature differentiation (against objective rivals - and Broadcom customers) and faster time to market. AGI (and Team Green) in sight? The small print of the photo showing OpenAI CEO Sam Altman and Broadcom CEO Hock Tan spelt out a surprising motto. "May we scale smoothly, exponentially and uneventfully through AGI". Jalapeño and its successors may well be instrumental to 'reaching' Artificial General Intelligence. The press release mentions a 'multi-generation' compute platform and I wouldn't be surprised if the future iterations notch up the heat with chilli names like Serrano, Cayenne, Habanero and Moruga. Tan told CNBC there will be a small prototype development by the end of 2026, adding, "We will start seeing it really ramp up in '27 and really going full tilt in first half '28". Whispers of the collaboration were first heard in July 2024 with an official announcement made at GITEX in October 2025. OpenAI has signed a flurry of strategic agreements with the likes of AMD, AWS and Cerebras to secure enough AI chips and reduce its dependency on Nvidia. Likewise, Broadcom has positioned itself as a key provider for anyone with deep enough pockets to yearn for their own hardware. Other than OpenAI, Google, Meta, and ByteDance are also likely customers with Anthropic, Apple and Fujitsu craving for Broadcom's expertise in ASIC and its custom AI accelerator portfolio. Becoming the kingmaker will help Broadcom close the gap between itself and Team Green, the nickname given to Nvidia. In a veiled dig towards the world's most valuable company, OpenAI said that its new platform was "not a general-purpose accelerator adapted from earlier AI workloads", a clear allusion to Nvidia GPUs that were used for training. The next big iteration of ChatGPT, GPT-5.6 is expected to be released this week. Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button! And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.

OpenAI tests homegrown AI chips

Why it matters: OpenAI is joining other leading AI companies in designing its own silicon as it races to secure more computing capacity, lower costs and reduce its dependence on Nvidia. Driving the news: This first generation of chips, which are aimed at inference rather than training, were developed with some help from Broadcom. * OpenAI did the core design, while Broadcom brought specific knowledge in connectivity and other areas. * OpenAI said its chips are specifically designed for handling current and future models, allowing them to be used more efficiently and deliver better performance per watt of electricity than off-the-shelf options. Zoom in: OpenAI is using the first sample chips in its labs for tasks similar to answering Codex queries. * They're delivering even better thermal performance than anticipated, the company says. * Broadcom said to expect the first chips to be in commercial use at Microsoft and other partners by the end of the year, though OpenAI says the real volume will come next year. * The company has said it aims to have the custom chips powering 10 gigawatts' worth of compute by 2029. What they're saying: Broadcom CEO Hock Tan said that the work with OpenAI highlights a message he has been pushing for a while: Companies that want to lead in AI need their own chips. * "At the end of the day, you cannot, should not rely on some other third-party GPU to do it for you, because it's such a key part," Tan said. * Richard Ho, who leads OpenAI's chip efforts, said that being able to bring AGI to benefit all of humanity requires the company to deliver compute efficiently and cost-effectively. * "This gives OpenAI full stack control," Ho said. The big picture: OpenAI has relied almost exclusively on Nvidia chips for both training and inference, though recently it began also using chips from Cerebras for inference. * Nvidia remains a key partner, especially for training new models, OpenAI says. * Google, Amazon, Microsoft and others already have their own AI chips that they use alongside Nvidia's processors. What we're watching: Whether OpenAI expands its homegrown chip use to training -- not just inference, something the company said it is considering.

OpenAI Turns Up the Heat With Jalapeño, Its First Custom AI Chip

The reveal follows months of reports that OpenAI was developing custom silicon to reduce reliance on Nvidia. OpenAI on Wednesday unveiled Jalapeño, its first custom artificial intelligence chip, giving a spicy name to a project that could reshape how the company powers ChatGPT and future AI products. Developed with Broadcom, the chip is designed specifically for large language model inference -- the process that generates responses to user prompts -- and marks OpenAI's most significant move yet toward controlling more of the hardware stack behind its AI models. "Jalapeño 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," OpenAI President Greg Brockman said in a statement. "By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access." Unlike many AI chips designed to handle a wide range of computing tasks, Jalapeño was built specifically to run chatbots and other AI systems powered by large language models. OpenAI said early versions of the chip are already being tested in its labs, including with GPT-5.3-Codex-Spark. The company claims Jalapeño can deliver more computing power while using less energy than today's leading AI chips, though it has not yet released benchmark results. The announcement confirms reports that OpenAI was building a custom chip program. Earlier this year, Reuters reported that the company was planning to launch its first internally deployed AI chip as it sought to reduce its reliance on Nvidia hardware. OpenAI's chip ambitions gained further traction in April, with reports that the company was developing a chip designed for smartphones. The chip is the first product in what OpenAI describes as a multi-generation compute platform expected to begin deployment in data centers later this year, with future generations supporting gigawatt-scale AI infrastructure with Microsoft and other partners. "Our collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI," Broadcom President and CEO Hock Tan said in a statement. "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."

OpenAI, Broadcom debut custom Jalapeño chip for AI inference

OpenAI, Broadcom debut custom Jalapeño chip for AI inference OpenAI Group PBC today revealed a custom chip called Jalapeño that it will use to power its large language models. The processor is the fruit of a collaboration with Broadcom Inc., which is no stranger to custom silicon design. The company helped Google LLC develop its TPU line of artificial intelligence accelerators. In April, the search giant extended its chip collaboration with Broadcom to 2031. Nvidia Corp.'s flagship Rubin graphics cards can run both training and inference workloads. By contrast, Jalapeño is only designed for the latter use case. According to OpenAI, early testing indicates that the chip can perform inference with significantly higher performance per watt than "current state-of-the-art," which may be a reference to Nvidia chips. The company has shared few details about Jalapeño's design. However, the blog post in which it announced the chip specifies that the underlying "architecture reduces data movement." That hints Jalapeño's architecture may be designed to reduce data movement between its logic circuits and off-chip memory, one of the main performance bottlenecks in inference clusters. AI chip suppliers take several approaches to reducing data movement. One of the most common methods is to equip an accelerator with a large amount of onboard SRAM, a type of high-speed memory. The more SRAM a chip includes, the less data must be sent to off-chip memory. Cerebras Systems Inc. and Groq Inc. are among the companies that have adopted that approach. OpenAI says that its Jalapeño-powered inference clusters will use multiple Broadcom networking technologies. One of them is the company's Tomahawk chip series, which is designed to power Ethernet switches. Tomahawk-based switches can be used to move data both between servers in the same rack and between racks. Broadcom's newest Tomahawk chip, the Tomahawk 6, can process up to 1.6 terabits of traffic per second. A built-in congestion management engine fixes network bottlenecks that might slow down connections. OpenAI plans to deploy Jalapeño and its Broadcom-supplied network equipment in custom server racks. The ChatGPT developer is developing the systems in collaboration with Celestia Inc., a Toronto-based provider of data center equipment design services. The company can also help customers optimize their server production lines. OpenAI will bring its first Jalapeño servers online by year's end. The company plans to expand its use of the chip over time. OpenAI's blog post describes Jalapeño as the "first step in a multi-generation compute platform," which hints that it may be planning to develop additional inference processors in the future. Another possibility is that OpenAI will design custom chips for adjacent use cases such as model training. Jalapeño may have the potential to open new revenue streams for the company. Nvidia sells its graphics cards as part of systems called DGX appliances that also include central processing units, cooling modules and other hardware. OpenAI has the resources to bring competing Jalapeño-powered appliances to market. It could even enable customers to run its AI models on-premises using such systems. A move into the lucrative AI hardware market might not only boost OpenAI's revenue growth but also raise investor interest in its upcoming public offering. Anthropic PBC, the company's top rival, recently filed for a listing of its own. An inference hardware offering could be a valuable differentiator for OpenAI during its roadshow, particularly if Anthropic goes public first.

OpenAI unveils first custom inference chip named Jalapeño

OpenAI revealed its inaugural custom-built inference processor, named Jalapeño, which was developed in collaboration with Broadcom. The processor caters specifically to the distinct requirements of OpenAI's inference systems, with the company stating that its own AI models contributed to its development. The chip is still undergoing testing, but early results indicate a significant improvement in performance-per-watt relative to current leading alternatives. OpenAI's partnership with Broadcom was officially announced in October, and the creation of custom chips has been viewed as a strategy to reduce reliance on Nvidia's graphics processing units. Google and Amazon have created comparable custom chips, termed "AI accelerators," to accelerate machine learning tasks. In an in-house podcast, OpenAI president Greg Brockman discussed the company's chip development strategy after announcing the partnership with Broadcom. "We have a deep understanding of the workload," Brockman said. "We've really been looking for specific workloads that are underserved, [and asking] how can we build something that will be able to accelerate what's possible?" Jalapeño is tailored for inference tasks, which involve the application of pre-built AI models based on user commands. According to OpenAI, the chip offers low operating costs when managing real-time coding models. However, performance-intensive processes such as pre-training may still require Nvidia hardware. Even minor reductions in inference costs could significantly enhance OpenAI's profitability. Optimizing the inference system is critical for the future economics of AI, and the company is expanding its capabilities across the entire technological stack. OpenAI is simultaneously developing products like Codex and the models that support them along with establishing data centers for model deployment. The shift to custom silicon is expected to further enhance these operational efficiencies. OpenAI detailed that its strategy encompasses the design of infrastructure components, including chip architecture, kernels, memory systems, networking, scheduling, and deployment systems. This comprehensive approach allows for optimization across all layers of technology, aiming to deliver faster, more reliable, and cost-effective models for users.

OpenAI's New 'Jalapeño' AI Chip Will Power Next-Generation AI

It promises superior performance per watt compared to other solutions OpenAI on Thursday unveiled its first custom-designed artificial intelligence (AI) accelerator called Jalapeño. It has been developed in collaboration with California-based semiconductor giant Broadcom. Instead of training, the San Francisco-based company describes it as an "Intelligence Processor" that is purpose-built for large language model (LLM) inference. It is claimed to deliver better performance per watt than current state-of-the-art solutions around compute, memory, and networking requirements. Jalapeño and OpenAI's Long-Term Infrastructure Strategy The AI firm said that Jalapeño has been designed to handle inference workloads powering its services like ChatGPT, Codex, the OpenAI API, and future agentic AI products. This sets it apart from conventional AI accelerators that have been adapted from general-purpose architectures. All aspects of the chip, including its architecture, board layouts, rack integration, networking, and deployment systems, have been developed in collaboration with Broadcom and manufacturing partner Celestica. As per the company, engineering samples of Jalapeño are already running machine learning (ML) workloads at production target frequencies and power levels in the company's labs, including GPT-5.3-Codex-Spark. OpenAI claims the new AI accelerator can reduce data movement while balancing compute, memory, and networking resources to achieve utilisation much closer to the hardware's theoretical limits. However, performance figure benchmarks have yet to be revealed by the San Francisco-based AI startup. Greg Brockman, President and Co-Founder of OpenAI, described Jalapeño as part of the company's effort to make AI faster, more reliable, and more affordable, stating that the world is rapidly moving towards a "compute-powered economy". The company expects Jalapeño to become the foundation of a multi-generation AI compute platform beginning next year. Broadcom and OpenAI both claim that Jalapeño moved from the initial design stage to manufacturing tape-out in nine months. This is described as one of the fastest ASIC development cycles achieved for a high-performance AI semiconductor. The former contributed its silicon implementation expertise and high-speed networking technologies, such as its Tomahawk networking silicon, while the latter will help deploy the platform across large-scale AI data centres. The introduction of the chip is a notable step in OpenAI's long-term infrastructure strategy of designing its own technology stack instead of relying entirely on third-party AI hardware.

OpenAI unveils custom chip it designed with Broadcom to boost its AI infrastructure

OpenAI has unveiled its first custom AI chip, Jalapeno, developed with Broadcom to boost its infrastructure. This move aims to overcome computing power shortages and reduce reliance on Nvidia's GPUs. Designed for AI inference, the chip is comparable to Nvidia's and Google's offerings. OpenAI plans to deploy Jalapeno by year-end, marking a significant step in its multi-generation chip development strategy. OpenAI showed off the company's first custom artificial intelligence chip designed in conjunction with Broadcom on Wednesday, as it seeks to speed its development of its infrastructure. AI labs such as OpenAI and Anthropic are struggling to obtain enough computing horsepower to run the latest, most powerful chatbots and coding apps. Some, such as OpenAI, have turned to developing in-house chips in order to reduce the cost and create an alternative to Nvidia's graphics processing ⁠units (GPUs) that ⁠are commonly used for AI. OpenAI's engineers designed the chip, called Jalapeno, together with Broadcom to perform a specific AI task known as inference, during which data is crunched in order to answer a user's query to a chatbot like ChatGPT. The chip made by the team is as good as the Blackwell chips made by Nvidia or the tensor processing units designed by Alphabet's Google, Broadcom CEO Hock Tan said in an interview with Reuters. The Jalapeno processor is designed to work speedily and efficiently with the large ⁠language models (LLMs) that power many AI applications, OpenAI hardware chief Richard Ho said. "It will be performant on, we think, all kind of future iterations of LLMs," Ho told Reuters. The company plans ⁠to deploy Jalapeno by the end of this year, and it is the first step in a multi-generation chip development plan, OpenAI said. Canadian electronics manufacturer Celestica will build the server systems, which, like the chips will be used only by OpenAI. The San Francisco company said it has samples of the chip running in its labs and they were operating at the target power and performance with the company's GPT-5.3-Codex-Spark AI model. It took the company's engineers roughly nine months to complete the chip design before it sent it to Taiwan's TSMC for manufacturing, in part because of using AI to speed specific aspects of the process, OpenAI said. Reuters first reported OpenAI was exploring making its own chip in 2023. To make their own in-house chips, Meta Platforms, Amazon and Google have turned ⁠to the likes of Broadcom and Marvell, which provide specific design services and intellectual property that can be difficult to replicate in-house. Anthropic is weighing building an AI chip of its own, sources told Reuters in April. At the moment, however, because of the AI-related surge in memory demand, Broadcom's profit margin on the custom chips is not as high as some of the other chips it makes, such as networking switches, Tan said. AI chips require large amounts of high-bandwidth memory, which challenges Broadcom's margins on custom AI chip products, Tan said. Tan said South Korea's SK Hynix and Samsung Electronics supply Broadcom with memory chips.

OpenAI's First Custom Chip Is As Hot As A Jalapeño For AI, As The Firm Calls It The "Best Inference Platform" for LLMs

OpenAI has just announced its first custom chip called Jalapeño, which is aimed at AI Inference workloads & produced by Broadcom. OpenAI Designed & Built Its First Custom AI Silicon, The Jalapeño Chip, For Agentic AI Today, OpenAI announced that it has built and designed its first AI chip, which they are calling Jalapeño. Interesting choice of name for a chip that came fresh off the baking ovens at Broadcom, who are responsible for producing this chip. The Jalapeño Intelligence Processor and its first wafer were shown off by OpenAI's CEO, Sam Altman, and Broadcom's CEO, Hock Tan. This chip continues the trend of AI firms developing custom AI chips for the Agentic AI era. We have seen the likes of Anthropic exploring custom AI chips, while Google is already deep within its TPU strategy with a multi-front solution. With the Jalapeño chip, OpenAI claims that it will mark the beginning of their vision of the future of LLM inference. The chip is designed to be OpenAI's first AI accelerator in a multi-generational compute platform that firms are building together to make AI faster, more reliable & accessible. Jalapeño is designed from scratch and focuses solely on AI workloads. OpenAI states that the chip was co-developed from initial design to manufacturing tape-out in just nine months. The chip will be backed by a robust ecosystem developed by its partners, Broadcom & Celestica. Together, the firms will help industrialize the platform through chip implementation, board, rack system integration, high-performance networking, and scalable production systems. Jalapeño is a blank-slate design for modern LLM inference, not a general-purpose accelerator adapted from earlier AI workloads. It is informed by the systems OpenAI runs every day across ChatGPT, Codex, the API, and future agentic products, while also being designed for current and future LLMs across the industry. The goal is to combine the power and throughput of today's leading AI accelerators with latency closer to the fastest specialized inference systems, making Jalapeño well suited for interactive LLM products at scale. OpenAI In terms of applications, Jalapeño is designed to be flexible and works with all LLMs. The first engineering samples of the chip are already running ML workloads such as GPT-5.3-Codex-Spark at production target frequency and power. The company isn't sharing a whole lot of details at the moment, but we do see eight HBM sites and the compute die(s) in the middle of the chip. As for launch, the first Jalapeño platforms are said to be deployed by the end of 2026 and will expand in the years ahead. The chip is designed to be a multi-generation effort with a multi-generation compute platform. Just like the others, OpenAI diving into the custom silicon space shows that the demand for custom ASIC and AI accelerators is growing a lot. Last year, OpenAI signed a partnership to deploy 10GW of NVIDIA systems, but as supply issues persist, AI firms are now investing in custom silicon that doesn't lock them into NVIDIA's ecosystem or make them solely reliant on one chipmaker, but makes their compute platform more diverse, with room to expand their compute portfolio. Follow Wccftech on Google to get more of our news coverage in your feeds.

OpenAI and Broadcom Launch Custom-built AI Chip Jalapeño to Accelerate AI Infrastructure Capabilities

Jalapeño was delivered to OpenAI CEO Sam Altman and President Greg Brockman by Broadcom President and CEO Hock Tan and President Charlie Kawwas, marking an important step in OpenAI's strategy to build the full stack behind its models and products. OpenAI and Broadcom unveiled Jalapeño, OpenAI's first Intelligence Processor: an accelerator architected around OpenAI's 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. Jalapeño was delivered to OpenAI CEO Sam Altman and President Greg Brockman by Broadcom President and CEO Hock Tan and President Charlie Kawwas, marking an important step in OpenAI's strategy to build the full stack behind its models and products. OpenAI designed the chip from scratch around its deep understanding of LLM fundamentals, informed by its roadmap of models, kernels, serving systems, and product needs, with partners Broadcom and Celestica, helping industrialize the platform through chip implementation, board, rack system integration, high-performance networking, and scalable production systems. Jalapeño is designed with flexibility to work with all LLMs guided by OpenAI's insights into the inference needs of current and future AI models across the industry. Engineering samples of the Jalapeño chip are running ML workloads in the lab at production target frequency and power, including GPT‑5.3‑Codex‑Spark. While OpenAI is still measuring final performance, early testing shows that Jalapeño will deliver performance per watt substantially better than current state-of-the-art. A detailed technical report on performance will be presented in the coming months. The architecture reduces data movement and balances compute, memory, and networking resources to achieve realized utilization much closer to theoretical peak performance. Broadcom's silicon implementation and networking technologies, including Tomahawk networking silicon, help bring the platform to large-scale production. "The world is moving to a compute-powered economy," said Greg Brockman, President and Co-Founder of OpenAI. "Jalapeño 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." "Jalapeño was designed from the ground up for LLM inference using detailed insights from our close collaboration with OpenAI researchers," said Richard Ho, who leads OpenAI's hardware program. "We optimized the architecture around the kernels, memory movement, networking, and serving patterns that matter most for frontier AI models. Based on early testing, Jalapeño will efficiently execute our most important workloads close to the hardware's theoretical limits." "Our collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI," said Hock Tan, President and CEO, Broadcom. "This 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." That is the full-stack advantage. OpenAI is not only developing frontier models or building products on top of them; it is designing the infrastructure underneath them: chip architecture, kernels, memory systems, networking, scheduling, deployment systems, and product experience. Because OpenAI operates across the stack, each layer can be optimized around the same goal: making its models faster, more reliable, and more affordable for users. Jalapeño strengthens the flywheel behind OpenAI's progress. Better infrastructure drives compute efficiency. Greater compute efficiency enables better training and serving, ultimately powering more capable AI models. Better models become better products for people, developers, and businesses. Better products drive more usage, more customers, and more revenue, which lets OpenAI reinvest in the next generation of infrastructure. Over time, that cycle helps make intelligence more capable, more reliable, and less expensive for everyone. The same models served to users are helping improve the infrastructure used to run future models. If AI can help engineers design better chips faster, it can lower the cost of compute across the industry and help democratize access to advanced AI. Democratizing AI means making advanced models available, dependable, and affordable enough for more people to use every day. Jalapeño helps OpenAI turn more of its infrastructure into useful intelligence for students, developers, small businesses, researchers, enterprises, and anyone trying to learn, create, or solve hard problems.

OpenAI and Broadcom unveil Jalapeño Intelligence Processor for LLM workloads

OpenAI and Broadcom have announced Jalapeño, OpenAI's first Intelligence Processor. The AI accelerator is designed around OpenAI's vision for the future of large language model (LLM) inference and is the first chip in a multi-generation compute platform being developed with Broadcom and Celestica. Jalapeño was delivered to OpenAI CEO Sam Altman and President Greg Brockman by Broadcom President and CEO Hock Tan and President Charlie Kawwas. OpenAI Jalapeño OpenAI designed Jalapeño from the ground up using its understanding of LLM architectures, model roadmaps, kernels, serving systems, and product requirements. Broadcom contributed silicon implementation, networking, and connectivity technologies, while Celestica provided board, rack, system integration, and production expertise. Jalapeño is designed to support current and future LLMs and is informed by the inference requirements of AI models across the industry. The chip architecture is based on systems OpenAI operates across ChatGPT, Codex, API services, and future agentic products. Engineering samples are already running machine learning workloads at target production frequency and power levels, including GPT-5.3-Codex-Spark. Key Features * Purpose-built architecture for modern LLM inference * Designed specifically for AI inference rather than adapted from earlier accelerator designs * Supports current and future large language models * Optimized around models, kernels, serving systems, and product requirements * Reduces data movement across the system * Balances compute, memory, and networking resources * Targets utilization closer to theoretical peak performance * Designed to deliver higher performance per watt than current state-of-the-art AI accelerators * Combines high throughput with latency closer to specialized inference systems * Supports interactive AI workloads at scale * Uses Broadcom's Tomahawk networking silicon * Supports large-scale deployment and production environments OpenAI said performance testing is ongoing and a detailed technical report will be released in the coming months. AI Infrastructure and Development OpenAI said Jalapeño is part of its infrastructure stack, which includes chip architecture, kernels, memory systems, networking, scheduling, deployment systems, and product experience. The chip was co-developed with Broadcom in nine months from initial design to manufacturing tape-out. The companies described it as the fastest ASIC development cycle achieved for a high-performance advanced semiconductor program. OpenAI and Broadcom attributed the development timeline to software-hardware co-development and the use of OpenAI models during parts of the design and optimization process. According to the companies, the same AI models used by customers today helped improve infrastructure that will support future AI systems. OpenAI said AI-assisted chip development could help reduce compute costs. Multi-Generation Platform Jalapeño is the first accelerator in a planned multi-generation compute platform scheduled for initial deployment by the end of 2026. The platform combines: * OpenAI-designed AI accelerators * Broadcom silicon implementation technologies * Broadcom networking and connectivity technologies * Celestica board, rack, and system integration expertise Additional generations of the platform are planned in the coming years. Benefits OpenAI said improvements in inference cost, speed, and reliability can contribute to: * Faster ChatGPT responses * Codex tasks that can perform more steps with less waiting * Lower-cost API products * More dependable access during periods of high demand The company said the infrastructure is aimed at supporting students, developers, small businesses, researchers, enterprises, and other users. Availability Engineering samples of Jalapeño are currently running production-target machine learning workloads in laboratory environments. Initial deployment of the Jalapeño platform is planned by the end of 2026. Speaking on the announcement, Greg Brockman, President and Co-Founder of OpenAI, said: The world is moving toward a compute-powered economy. Jalapeño is part of our long-term full-stack infrastructure strategy to make compute more abundant, helping deliver AI that is faster, more reliable, more affordable for people and businesses, and capable of addressing more important problems. By designing more of the stack ourselves, we can provide more intelligence with greater efficiency and continue advancing AI toward broader access. Commenting on the development, Hock Tan, President and CEO of Broadcom, said: Our collaboration with OpenAI reflects a long-term commitment to scaling the physical infrastructure needed for the next decade of AI. This marks the beginning of a multi-generation roadmap. By co-developing our industry-leading silicon with OpenAI, we are supporting the deployment of gigawatt-scale data centers with Microsoft and other partners starting in 2026.

OpenAI Unveils Custom Chip Co-Designed with Broadcom to Boost AI Infrastructure

OpenAI may have finalised deals worth billions of dollars with chipmakers ranging from Nvidia to AMD and Amazon, but the structural challenge it faced to its AI dreams finally seems to have finally paid off. The company joined hands with Broadcom to showcase its first custom AI accelerator chip named Jalapeno. This move puts the Sam Altman-led company a step ahead of its arch rival Anthropic, since both have been racing each other to generate compute horsepower to run their chatbots and coding apps. Also worth noting would be how the future would pan out for OpenAI, now that it has an in-house processor, especially with chipmaking giants like Nvidia, AMD and Cerebras. OpenAI hardware chief Richard Ho clarified that their engineers designed the chip to assist in LLM-based inference where massive volumes of data is crunched for answering a user's single query. Broadcom CEO Hock Tan told Reuters that the chip is as good as Blackwell chips designed by Nvidia or Google's tensor processing unit (TPU).

OpenAI and Broadcom unveil Jalapeño AI inference chip By Investing.com

SAN FRANCISCO and PALO ALTO, Calif. - OpenAI and Broadcom Inc. (NASDAQ:AVGO) announced today the release of Jalapeño, OpenAI's first custom AI accelerator chip designed for large language model inference, according to a press release statement. The partnership marks a significant win for Broadcom, a prominent player in the semiconductor industry with a market capitalization of $1.81 trillion and revenue growth of 32% over the last twelve months. The chip was developed from design to production in nine months through collaboration between OpenAI, Broadcom, and Celestica. Broadcom handled chip implementation and networking technologies, while Celestica contributed board and system integration expertise. Jalapeño is designed specifically for LLM inference workloads rather than as a general-purpose accelerator. Engineering samples are currently running machine learning workloads in laboratory testing at production target frequency and power levels, including GPT-5.3-Codex-Spark. Early testing indicates the chip will deliver performance per watt better than current state-of-the-art solutions, though OpenAI stated it is still measuring final performance metrics. A detailed technical report on performance will be released in coming months. "Jalapeño 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," said Greg Brockman, President and Co-Founder of OpenAI. The chip incorporates Broadcom's Tomahawk networking silicon and is designed to work with LLMs across the industry. The architecture aims to reduce data movement and balance compute, memory, and networking resources. Broadcom President and CEO Hock Tan said the collaboration represents "a fundamental commitment to scaling the physical infrastructure required for the next decade of AI" and noted plans for deployment at gigawatt scale data centers with Microsoft and other partners beginning in 2026.The announcement comes as Broadcom trades at $380.15, which InvestingPro analysis suggests is undervalued relative to its Fair Value. With 25 analysts recently revising earnings upwards and the stock delivering a 45% return over the past year, investor confidence remains strong. The company maintains impressive gross profit margins of 76%, according to InvestingPro data, which tracks over 1,400 US equities with comprehensive Pro Research Reports that transform complex financial data into actionable intelligence. The companies described this as the first step in a multi-generation compute platform designed for initial deployment by the end of 2026. OpenAI used its own AI models to accelerate parts of the chip design and optimization process during development. In other recent news, Broadcom Inc. has announced the pricing terms for a $2.5 billion cash tender offer to purchase outstanding senior notes. The notes, with maturities ranging from 2030 to 2038, include several series with varying interest rates. This financial maneuver is part of Broadcom's strategy to manage its debt portfolio. Meanwhile, JPMorgan has reiterated an Overweight rating for Broadcom, maintaining a $580.00 price target, and dismissing concerns about potential delays in the company's TPU v9 2nm ASIC program. The firm assures that Broadcom remains on schedule for a 2028 ramp-up of this next-generation chip. Additionally, Wolfe Research has highlighted Broadcom's discussions on financing vehicles, which could significantly impact future revenue from XPU shipments. UBS has also reiterated a Buy rating, citing Broadcom's AI infrastructure deals and a 20GW special purpose vehicle agreement with Apollo and Blackstone for OpenAI and Anthropic. These developments indicate ongoing strategic initiatives and analyst confidence in Broadcom's future performance. This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.

OpenAI, Broadcom Develop Custom Chip for AI Inference

OpenAI and Broadcom collaborated to develop a custom artificial-intelligence chip, marking the next step in the ChatGPT-parent's strategy to build out a full stack behind its models and products. The chip, called Jalapeño, was built from scratch specifically for large-language model inference, with the goal of improving efficiency and lowering the cost to serve AI systems at scale, the companies said Wednesday. Early tests show that the chip will deliver performance per watt "substantially better" than current state-of-the-art chips, they added. OpenAI and Broadcom didn't disclose specific performance measures but said a more-detailed technical report will be presented in the coming months. Jalapeño was built around the memory, networking and compute patterns of large language models, rather than adapted from general-purpose architectures, according to the companies. It also reflects OpenAI's broader effort to reduce reliance on external hardware suppliers and optimize infrastructure for its growing AI services. "The world is moving to a compute-powered economy," OpenAI President and Co-Founder Greg Brockman said. "By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access." Broadcom, which is providing silicon implementation and networking technology for the effort, said the project represents the start of a multi-generation collaboration aimed at building large-scale AI infrastructure. The collaboration comes after OpenAI and Broadcom said last fall they were working together to develop and deploy 10 gigawatts of custom AI chips and computing systems over the next four years. The move places OpenAI among a growing list of technology companies that are designing their own chips to better control the cost and performance of AI workloads.

OpenAI and Broadcom's Jalapeño: What this spicy AI chip means for NVIDIA, Qualcomm and Google

For many years now, NVIDIA has maintained an almost uncontested monopoly in the AI infrastructure market. The H100 and B200 GPUs are used to run data centers that host ChatGPT, Gemini, and all other advanced AI models that are being used. However, there is news from OpenAI which hints at a changing landscape - the Jalapeño, which is an accelerator chip customized for large language model inference in collaboration with Broadcom. Also read: Claude Tag explained: The AI tool that writes 65% of Anthropic's code The Jalapeño chip was announced by OpenAI on June 24, 2026. It is the first Intelligence Processor that OpenAI has designed, and this means that it is a custom-built accelerator chip based on the requirements of the company's vision for the future of LLM inference. This is not a repurposed general-purpose GPU; rather, it is a chip designed with the needs of modern LLM inference in mind. This is important because the actual performance is going to be much better than what the current generation has to offer. The architecture ensures that the performance-per-watt is far better than what exists right now with some more performance numbers expected to come in the near future. The architecture, according to OpenAI, has been designed such that data transfers are minimal and computing, memory, and networking elements have been balanced to get the maximum utilisation close to the peak performance - something that is known to be difficult for general-purpose GPUs when dealing with LLM workloads. Also read: WhatsApp under Will Cathcart: The controversies that defined its last seven years What makes Jalapeño stand out apart from its specs is the speed with which it was developed. From design to manufacturing tape-out, the chip was done in nine months flat and OpenAI boasts that it's the fastest development time for an ASIC till date in high-performance semiconductors. This feat has been made possible partly due to the use of their models in certain areas of the design process. So what does this mean for the competition? This marks yet another chip-maker joining the list of customers who have gone in-house. There are Google's TPUs, Amazon's Trainium, Microsoft's Maia, as well as the custom silicon for inference developed by Apple and Qualcomm. In essence, there is a threshold level of scale beyond which developing your own silicon makes sense from both an economic and strategic standpoint. With ChatGPT being used by hundreds of millions of people, the potential savings are tremendous. For Qualcomm, this move means less than for other companies mentioned above. Qualcomm has been actively promoting its Snapdragon and cloud AI chips for inference on devices and at edge locations. Jalapeño is an accelerator targeting the data center and not an edge rival. However, it shows that OpenAI aims to take control of the entire stack for inference, and such ambitions may eventually extend even beyond the data centers. For Google, which runs its own TPU infrastructure and directly competes with OpenAI on the market, Jalapeño reduces the difference in terms of self-sufficiency of the infrastructure. The unique advantage of Google was in controlling the hardware for its AI. Jalapeño is planned for initial deployment by end of 2026, at gigawatt scale with data centre partners, as the first step in a multi-generation platform. The chip wars just got a new, very spicy contender.

OpenAI introduces its first AI chip designed for LLM workloads: Check details

OpenAI and Broadcom developed Jalapeno from the initial design to manufacturing tape-out in just nine months. OpenAI has officially introduced Jalapeno, its first custom AI chip built for large language models (LLMs) workloads. The chip has been developed in partnership with Broadcom and marks a major step in OpenAI's effort to build more of the technology behind ChatGPT and its other AI products. According to the company, Jalapeno is not a general-purpose AI chip. Instead, it has been designed from the ground up for AI inference. OpenAI describes Jalapeno as its "first Intelligence Processor: an accelerator architected around OpenAI's 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." Also read: GTA 6 pre orders now live: India price, benefits and other details OpenAI said engineering samples of Jalapeno are already running AI workloads in its labs, including GPT-5.3-Codex-Spark. While the company is still testing the final performance, early results show that the chip offers significantly better performance per watt than current state-of-the-art. A detailed technical report will be shared in the coming months. "Jalapeno 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," said Greg Brockman, President and Co-Founder of OpenAI. "By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access." One of the biggest highlights is the speed of development. OpenAI and Broadcom developed Jalapeno from the initial design to manufacturing tape-out in just nine months. The companies believe this is the fastest ASIC development cycle ever achieved for advanced high-performance semiconductor chips. Also read: Anthropic launches Claude Tag, an AI agent designed to collaborate across organisations: How it works "If AI can help engineers design better chips faster, it can lower the cost of compute across the industry and help democratize access to advanced AI," OpenAI said. According to OpenAI, Jalapeno is the "first step in a multi-generation compute platform." The company plans to start deploying the new platform by the end of 2026. "The point of this work is simple: inference is where AI reaches people. Every improvement in cost, speed, and reliability can show up as a faster ChatGPT answer, a Codex task that can take more steps with less waiting, an API product that is cheaper to build, or more dependable access when demand is high," the company said.