Apple Collaborates with Broadcom on Custom AI Server Chip for Enhanced Cloud Processing

Apple may be working on a custom AI server chip with Broadcom's help

The big picture: Apple's possible foray into custom AI server processors reflects a broader trend in the tech industry. Companies like Amazon, Meta and Microsoft have already made significant strides in this area. Meanwhile, Google has also been leveraging Broadcom's intellectual property in its Tensor Processing Units. Apple is reportedly developing a custom server processor to power its AI services. Codenamed "Project Baltra," the initiative aims to bolster the AI capabilities integrated into Apple's operating systems, with production expected to begin in 2026, according to The Information, which cites three unnamed sources familiar with the matter. These sources indicate that Apple is partnering with semiconductor giant Broadcom for this endeavor. Apple now possesses a strong history and experience designing its own Arm-based silicon and already maintains an existing relationship with Broadcom in 5G component development. While specific details remain scarce, it is speculated that Broadcom's recent unveiling of its 3.5D eXtreme Dimension System in Package (3.5D XDSiP) technology could play a role in the project's development. This advanced packaging solution offers a multi-die processor architecture, which includes stacked compute dies on logic dies. It also integrates a high-bandwidth memory (HBM) interface and employs a face-to-face chiplet design to boost performance. The 2026 production target for Project Baltra aligns with Broadcom's projections for the production timeline of its 3.5D XDSiP technology. Apple could leverage this technology in several ways. The ability to integrate multiple dies and HBM stacks could allow Apple to create more powerful AI processors with higher compute density and improved efficiency. The 3.5D XDSiP platform also offers flexibility in design, enabling chip designers to pair different fabrication processes for each component, potentially allowing Apple to optimize various parts of its AI processor for specific tasks. The scalability offered by 3.5D XDSiP could help Apple design highly efficient AI processors to meet the demands of large language models and other compute-intensive AI workloads. As Moore's Law scaling becomes increasingly challenging, advanced packaging technologies like 3.5D XDSiP could allow Apple to push the boundaries of AI chip performance beyond the limitations of conventional process node improvements. In related news, Broadcom CEO Hock Tan recently projected the AI semiconductor market to reach $60 - $90 billion by 2027, emphasizing strategic partnerships in AI as key drivers of sustained growth. Following the company's latest quarterly report, Broadcom's stock surged over 20%, pushing its market capitalization above $1 trillion, making it the 10th largest company in the world. As with many of Apple's projects, details about Project Baltra are likely to remain under wraps until an official announcement. However, this initiative underscores Apple's commitment to advancing its AI capabilities. At a recent developer conference, Craig Federighi, Apple's Senior Vice President of Software Engineering, described the company's vision for "Apple Intelligence" to function both on-device and in a private cloud powered by Apple Silicon.

Apple is reportedly making an AI server chip with Broadcom's help

Apple appears to be developing the chips to enable the evolution of Apple Intelligence. The future of technology involves AI and Apple is doing what it can to be part of that future. To that end, a new report states that the company has a partnership in place to develop its AI hardware. According to The Information (subscription required), Apple is working with Broadcom on a server chip that can handle the demands of AI processing. The chip, code-named Baltra, is expected to be ready by 2026. While Apple is working with Broadcom on the development of Baltra, The Information reports that the actual manufacturing will involve TSMC's N3P process, an enhancement of 3nm production technology. Last May, Bloomberg reported on project ACDC (Apple Chips in Data Centers), Apple's effort to build a cloud infrastructure using its own chips. Project ACDC began in 2021, but with the AI boom, Apple prioritized the project in order to support its Apple Intelligence features. ACDC involved the use of M2 Ultra chips with plans for the M4; it's not clear if Baltra will be part of the ACDC project, but it does seem like a fit. AI technology is processor-intensive, so much so that many tasks are processed in the cloud-AI technologies from other companies mostly rely on cloud processing. Apple has been able to develop many Apple Intelligence features so that the processing occurs on the device, but as the features evolve they become more complex and require the processing power that the cloud provides. If Apple does successfully execute Baltra by 2026, that could be followed by a wave of more robust and complex Apple Intelligence features. AI chips are all the rage, with Google announcing its own Willow quantum AI chip just this week. Google says the chip, which represents the culmination of a decade of development, performed "a standard benchmark computation in under five minutes that would take one of today's fastest supercomputers 10 septillion years." On Wednesday, Apple introduced a second wave of Apple Intelligence features that include the Image Playground app where you can create AI-generated images; Genmoji, custom emoji that you create; ChatGPT-based writing tools, and more. The new features are available in the iOS 18.2, iPadOS, 18.2, and macOS Sequoia 15.2 updates. The Apple/Broadcom AI deal is the latest in the two company's shared history. Apple is working with Broadcom on RF filters for Apple's own 5G modem that could debut in the iPhone SE in 2025. Furthermore, a report last October stated that Apple is developing its own Wi-Fi chip that would replace a chip made by Broadcom.

Apple is reportedly building an AI server chip with Broadcom

Something called 'Baltra' expected to make its debut in 2026, perhaps with tech both already use Apple is reportedly working with chip giant Broadcom to develop a custom server processor to power the AI services and features built into its operating systems. According to The Information, which cites three unnamed sources with knowledge of the matter, the project is codenamed "Baltra" and is expected to enter production in 2026. Beyond this, details are predictably pretty thin. At the iGiant's developer conference earlier this year, Craig Federighi, SVP of software engineering, declared that Apple Intelligence would run both on-device and in a private cloud using servers powered by Apple Silicon. The idea that Apple might be building custom chips for GenAI isn't surprising. The iBiz has designed its own Arm-based silicon for years. The idea that Broadcom might be involved in this process shouldn't come as a shock either as the two companies already work together on 5G componentry. Broadcom is a massive conglomerate that, among other things, licenses intellectual property for use in semiconductor designs - particularly in the field of high-speed networking. For instance, back at the Hot Chips conference we looked at an optical interconnect chiplet designed to be packaged alongside GPUs and other accelerators to support larger scale-up compute clusters. More recently, Broadcom showed off its 3.5D packaging tech, which is designed to help chipmakers scale beyond reticle limits - similar to what Intel has done with its Ponte Vecchio (GPU Max) GPUs. AMD used similar techniques AMD to build its MI300X accelerators, which combine eight compute chiplets and vertically stack them on top of four I/O dies which handle memory management and chip-to-chip communication. Broadcom's 3.5D eXtreme Dimension System in Package tech - 3.5D XDSiP for short - is essentially a blueprint that customers can use to build multi-die processors of their own. Like AMD's MI300X it stacks compute dies on top of a logic die that interfaces with high bandwidth memory (HBM), but breaks out other I/O functionality into a separate set of dies. Broadcom's approach to interfacing compute dies interface with the rest of the system logic is a little different. According to Broadcom, previous 3.5D packaging technologies have used face-to-back approaches, which require more work to route through silicon vias (TSVs) which shuttle data between the two. Broadcom's design uses a face-to-face approach, which allows for denser electrical interfaces between the chiplets using hybrid copper bonding (HBC). We're told this will allow for substantially higher die-to-die interconnect speeds and shorter signal routing. The largest of these designs will support two 3D stacks, a pair of I/O chiplets, and up to 12 HBM3 modules on a single package totaling more than 6,000 mm of silicon area. Broadcom expects the first parts based on these designs to enter production in 2026. Coincidentally, that's the same timeline that The Information reports Apple's project Baltra is aiming at. It is unclear if the two projects are related. However some Apple chip designs - the M2 Ultra for instance - already make use of multi-die architectures, so it's not a stretch to suggest some overlap. We likely will not know much more about Baltra until it becomes reality. Apple is notoriously tight-lipped about products before they're officially announced. And while Broadcom is always happy to discuss its chip tech, it's highly secretive about who's actually buying it. Apparently that list includes some pretty big names - for instance, Google reportedly made extensive use of Broadcom IP in its Tensor Processing Units (TPUs). It's not a stretch to imagine Apple taking advantage of existing technologies rather than reinventing the wheel. The Register sought comment from Broadcom and Apple. Unsurprisingly, neither had responded at the time of publication. ®

Apple reportedly developing AI server chip with Broadcom | TechCrunch

Apple is working with semiconductor company Broadcom on its first server chip designed to handle AI applications, according to The Information, which cited three people with knowledge of the project. Apple is known for designing its own chips - called Apple Silicon and primarily manufactured by TSMC - for its devices. But those chips weren't necessarily designed to power AI processing. Apple on Wednesday rolled out several much-awaited Apple Intelligence features to users with newer iPhones and Macs, like its ChatGPT integration. While Apple plans to run much of its AI directly on devices, certain tasks - like Siri and Maps - are handled in the cloud and have heavy computing demands. And Apple plans to roll out even more generative AI features in the coming years. That's where Apple's new server chips, code-named internally Baltra, come into play. A source told The Information that Broadcom and Apple are focused on the chip's networking technology, which is critical for connecting a device to a network for AI processing. The company aims to complete the chip design within 12 months.

Apple Is Reportedly Working on Its First AI Server Chip With Broadcom

Apple and Broadcom could complete the chip's design in 12 months Apple is developing a new server chip that is designed to enable support for features powered by artificial intelligence (AI) in partnership with semiconductor manufacturer Broadcom, according to a report. The company previously announced that it would offload processing for some of its Apple Intelligence features to the cloud, but it has yet to utilise its own processor tailored for AI applications. Apple has also rolled out support for additional Apple Intelligence features on iOS, iPadOS and Mac computers, including ChatGPT integration with Siri. Citing three persons familiar with the company's plans, The Information reports that Apple is working on a server chip for AI applications. The iPhone maker already produces its own processors for its devices, which offer some on-device AI features, and the new in-house chip could be used to perform AI processing on Apple's servers. Apple's new server chip for AI processing is codenamed Baltra, according to the publication, and the companies are said to be working on the networking technology used by the processor, which will be used to perform AI tasks requested by a user on the cloud. The company's latest iOS 18.2, iPadOS 18.2, and macOS 15.2 updates -- released on Wednesday -- introduced new on-device Apple Intelligence features like Genmoji and Image Playground, as well as ChatGPT integration which requires access to the internet. The report states that the chip design is likely to be completed within a year, but it is currently unclear whether it will be used to power the company's existing AI features via Apple's Private Cloud Compute (PCC) cloud-based AI processing system that was announced earlier this year. With PCC, the company says it can offer support for Apple Intelligence features that rely on much larger server-based models -- these currently run on Apple Silicon chips -- to perform AI tasks that are too complex for on-device processing. Earlier this year, Apple stated that it would not store user data on PCC servers, except for handling a request and that users would be able to verify the company's privacy claims.

Report claims Apple is working with Broadcom to develop a custom chipset for AI - SiliconANGLE

Report claims Apple is working with Broadcom to develop a custom chipset for AI iPhone maker Apple Inc. is said to be working with Broadcom Inc. on the development of a customized server processor that will power the artificial intelligence features built into its iOS operating system. A report from The Information cites three unnamed sources as saying that the project has been given the codename of "Baltra", and the chip in question is slated to launch in 2026. Apart from that, the report provides few other details, which is not surprising given the secretive way both companies operate. What we do know is that, during Apple's developer conference earlier this year, the iPhone maker's senior vice president of software engineering, Craig Federighi, said some Apple Intelligence features will be powered by on-device chips, with others running on cloud-based servers that use the company's proprietary chips. If Apple is indeed developing its own chips for generative AI, it would not be much of a surprise. The company has been using its own, Arm-based central processing units for years already. It's also not a surprise to hear that Broadcom is working with Apple on such a project, as it already helps the company out with some of its 5G modems. It remains to be seen why Apple is tapping Broadcom's expertise again, but one theory is that it's interested in the company's semiconductor interconnect designs, which can enable faster chip-to-chip communications. Earlier this year, Broadcom showed off a new optical interconnect chiplet that's designed to work with graphics processing units and other types of AI accelerators, and it also unveiled its 3.5D packaging technology, which can help chips to scale beyond existing limits. Broadcom's 3.5D eXtreme Dimension System in Package, known as 3.5D XDSiP, is a kind of blueprint that customers can use to build multi-die processors similar to those found in Advanced Micro Devices Inc.'s MI300X GPUs. Those chips combine eight compute chiplets, vertically stacked on top of four input/output dies that handle chip-to-chip communications and memory management. Rather than stacking its chips vertically, Broadcom's design is said to use a face-to-face approach, enabling denser electrical interfaces between the chiplets, based on hybrid copper bonding or HBC technology. Broadcom has said this will deliver significantly faster die-to-die interconnect speeds and shorter signal routing, which translates to much higher bandwidth for chips to talk to one another. Broadcom said the largest of its 3.5D XDSiP designs can support two 3D stacks, a couple of I/O chiplets and as many as 12 High Bandwidth Memory or HBM modules on a single package. The company has said the 3.5D XDSiP packaging technology will enter production sometime in 2026. Given that Apple's project Baltra has a similar timeline, it therefore wouldn't be surprising if Apple's customer server processor is using the same technology, though that cannot be confirmed. However, there is a precedent, considering that some of Apple's chips, such as the M2 Ultra, already utilize multi-die architectures. For now this is all just speculation, and we probably won't know much more about Baltra until it's officially announced, which means probably not before 2026. That's because Apple is notoriously secretive about the new things it's working on, while Broadcom is just as notoriously secretive about who is using its technology.

Apple Is Working on AI Chip With Broadcom, Information Reports

(Bloomberg) -- Apple Inc. is developing a server chip designed especially for artificial intelligence and is working with Broadcom Inc. on the chip's networking technology, the Information reported, citing a person with direct knowledge of the project. The new chip is internally code-named Baltra and expected to be ready for mass production by 2026, according to the report. The move highlights Apple's longstanding position to avoid buying chips from Nvidia Corp., which dominates the market for AI processors. It also marks a milestone for the company's chips team, which began designing chips for the iPhone and moving on to Mac processors that set new standards for performance and energy efficiency, according to the report. Apple has been been playing catch up in the race for artificial intelligence. In October, the company began slowly introducing its first generative AI features, collectively known as Apple Intelligence, for the iPhone, iPad and Mac computers. Though Apple previewed a broader set of AI capabilities in June, the initial features represent only a sliver of its plans for the service. Apple will deliver some of its upcoming artificial intelligence features this year via data centers equipped with its own in-house processors, part of a sweeping effort to infuse its devices with AI capabilities. Bloomberg has previously reported that Apple is placing high-end chips -- similar to ones it designed for the Mac -- in cloud-computing servers designed to process the most advanced AI tasks for Apple devices.

Apple partners with Broadcom to make its server AI processors

To power its Apple Intelligence servers, Apple is now said to be at least in part working with Broadcom on a new processor design. Apple makes much of how Apple Intelligence works on-device, but it's also clear that there are requests that need use of a greater Large Language Model. Such requests are sent to Apple's servers, and the company has been working to produce server processors specifically for this purpose. According to The Information, that work includes a partnership with Broadcom to make a processor codenamed Baltra for release in 2026. Reportedly, this does not mean Broadcom designing and producing the new processor, but possibly only providing one of several "chiplets.". A processor's functions could be split up into several chiplets which Apple later recombines into a single chip. That would reportedly reduce manufacturing complexity, but also mean Apple could keep its overall design confidential, even from partner firms. It's not certain why Apple would want to partner with Broadcom when it is already developing its own Apple Silicon processors. However, AI servers may need many processors working in tandem, and Broadcom could be working on the networking between them. Reportedly, Apple cancelled the development of an unknown high-performance processor for Macs, in order to divert engineers to this AI server chip. The engineers in question are based in Israel. It's claimed that those engineers were instrumental in the move to Apple Silicon. Baltra, named after one of the Galapagos islands, is expected to be manufactured by TSMC using its N3P process. Announced in April 2024, the technology is expected to be first seen in the processors for the iPhone 17 Pro. Separately, it's previously been reported that Apple is in talks with Foxconn to build Apple Intelligence servers in Taiwan.

Apple Reportedly Working With Broadcom to Develop New AI Chip

Apple aims to support AI features on its devices with in-house server chips, continuing its success with M-series processors. Apple is reportedly working with Broadcom to develop its first server chip specifically designed for artificial intelligence (AI) processing, The Information reported on Wednesday, citing sources familiar with the matter. This move positions Apple alongside other tech giants that have developed proprietary chips to handle AI-intensive workloads and reduce dependence on Nvidia's costly and short-in-supply processors. Also Read: OpenAI is Developing In-House AI Chip with Broadcom and TSMC: Report Internally code-named Baltra, Apple's AI chip is slated for mass production by 2026. The company plans to manufacture the chip using Taiwan Semiconductor Manufacturing Co.'s (TSMC) advanced N3P process, according to the report. In 2022, Apple signed a multi-billion-dollar agreement with Broadcom to develop 5G radio frequency components. Earlier this year, at its annual developer conference, Apple revealed plans to leverage in-house server chips to support AI features across its devices. Also Read: Meta Announces USD 10 Billion AI Data Center in Louisiana Apple has made significant progress in chip development, notably with its M-series processors, which replaced Intel chips in Mac laptops. However, like many major tech firms, Apple faces challenges in reducing reliance on Nvidia, a hurdle that Google has managed to overcome by also partnering with Broadcom for its AI chips. Broadcom's primary competitor in this domain is Marvell. Marvell''s Chief Operations Officer Chris Koopmans, earlier this month said the custom chip market is projected to reach USD 45 billion by 2028, with Broadcom and Marvell expected to dominate the sector, according to a Reuters report.

Apple Is Working on AI Chip With Broadcom, Information Says

Apple Inc. is developing a server chip designed especially for artificial intelligence and is working with Broadcom Inc. on the chip's networking technology, the Information reported, citing a person with direct knowledge of the project. The new chip is internally code-named Baltra and expected to be ready for mass production by 2026, according to the report.

Apple Partners with Broadcom to Develop First AI Server Chips, Launch Set for 2026 - Decrypt

Apple is teaming up with semiconductor developer Broadcom to launch its first server chips dedicated to artificial intelligence. It follows Apple's release on Wednesday of its latest updates to iOS, macOS, and other devices featuring new Apple Intelligence capabilities. Apple and Broadcom did not immediately respond to Decrypt's request for comments. Apple is collaborating with Broadcom on the chip's networking features, according to a report by The Information on Wednesday, citing sources with "direct knowledge of the project." Code-named Baltra, the AI-focused chip is expected to enter production in 2026. Experts suggest that Apple's collaboration with Broadcom aims to bring the server needs for its growing Apple Intelligence initiatives closer to home. "Those capabilities can actually be processed on the phone, the iPad, the Mac, independently, but in some cases, they need to go to the cloud, things like writing tools, things like summarizing text, in some cases, summarizing articles," Bloomberg's Mark Gurman told Bloomberg TV. "And for that, you need your own server infrastructure." Founded in 1991, San Jose-based Broadcom develops semiconductor and software technologies. Its products are used in data centers, networking, and wireless communication, supporting various modern tech applications. This new collaboration is the latest in a partnership between Apple and Broadcom that began in 2023 with a multi-year deal for the chip manufacturer to develop 5G radio frequency components for Apple products. The new AI chip aims to meet the significant computing demands of Apple's enhanced AI features while improving performance and energy efficiency. Apple's Israel-based silicon design team will spearhead the development of the chip. Earlier this month, Apple acknowledged that its current AI-related computer chips are made through partnerships with Google Cloud and Amazon Web Services to support its search and Apple Intelligence programs. Others working with Apple to develop the new AI chip include Taiwan Semiconductor Manufacturing Company, which will be responsible for mass-producing the chip. Apple's AAPL stock rose briefly on Wednesday following the announcement of the partnership with Broadcom to $250 before falling to $247. Meanwhile, Broadcom's AVGO stock rose 7.5% to $184.73 before slipping to $182.98 in after-hours trading.

Apple Teams up with Broadcom for AI Chips, To Sideline NVIDIA | AI News

The chip, internally code-named "Baltra", is projected to enter mass production by 2026. Apple is collaborating with Broadcom to develop its first AI-focused server chip, a move aimed at preparing to address the intense computing demands of its AI features, according to a report by The Information. The chip, internally code-named "Baltra", is projected to enter mass production by 2026. The initiative mirrors strategies employed by other tech giants like Google, Amazon, and Meta, which have designed custom chips to support AI services. The collaboration highlights Apple's strategic effort to reduce dependence on NVIDIA, which dominates the AI chip market. The chip will reportedly use Broadcom's networking technology to optimise AI processing and employ advanced manufacturing processes from TSMC. Broadcom shares rose by five percent following the announcement, reflecting the company's growing role in AI chip manufacturing. TSMC's advanced N3P fabrication process is reported to be used for Baltra. Apple and Broadcom declined to comment on the report. Apple's in-house silicon team in Israel is leading the project, focusing on a chiplet design that improves efficiency and reduces complexity. The design incorporates Apple's Neural Engine technology to enhance AI task performance. This initiative aligns with Apple's recent AI advancements, including features that generate text, create images, and summarise content. While Apple's current chips power some AI features, they are not as efficient as chips designed specifically for AI processing. Broadcom, which has also partnered with Google on AI chips, will provide networking technology but allow Apple to retain control over the chip's overall design and production. This partnership also builds on an existing multi-billion-dollar deal between Apple and Broadcom to produce 5G radio frequency components, announced last year. Apple previously shared plans to integrate its own server chips to power AI features on its devices. Apple has achieved notable success with in-house chip development, such as the M-series processors, which replaced Intel chips in Mac computers. Despite such advancements, NVIDIA remains a dominant player in the AI chip market, though Google has managed to collaborate with Broadcom to lessen its dependency. In terms of collaboration, even OpenAI, the creator of ChatGPT, has partnered with Broadcom and TSMC to launch its first in-house AI chip in 2026. The generative AI boom has significantly boosted Broadcom's market presence, with its shares surging 54 percent in 2024. Analysts predict the custom chip market could grow to $45 billion by 2028, with Broadcom and competitor Marvell poised to share the expanding market. Apple plans to use a chiplet design for its AI chip, a concept first introduced by AMD over a decade ago, according to two sources. Instead of creating one large chip with different functional sections, Apple will split the chip into smaller, specialised pieces called chiplets, which will then be reassembled into a single unit. This approach simplifies manufacturing and reduces the chances of defects. However, NVIDIA remains a key player in tasks requiring heavy computing, such as training AI models. Apple's project is part of a broader strategy to integrate custom chips for secure, efficient AI data processing. As the company rolls out more AI features, the Baltra chip could play a crucial role in scaling services to billions of devices.

Apple Intelligence Servers Could Be Treated To A New AI Chip That Will Be Developed With A Helping Hand From Broadcom

To push into the generative AI space with unrelenting force, Apple Intelligence servers that were currently equipped with the M2 Ultra will eventually be replaced by the M4 Ultra, assuming the company releases it during the intended launch timeframe. However, it appears that a new chip is being developed by Apple, but the company is not taking this route on its own because a new report states that Broadcom will also be involved. A paywalled report published on The Information that was spotted by AppleInsider talks about the new AI chip, which could help boost the capabilities of Apple Intelligence servers by bringing in more cloud-based features. While the Cupertino firm's Large Language Models are being trained on Amazon's custom chips, there is also a need to address the ever-increasing demand of requests that will be handled by Apple Intelligence servers. This is where the involvement of Broadcom likely comes into play, with the new AI chip codenamed 'Baltra' and slated to launch in 2026. Baltra is said to leverage TSMC's 3nm 'N3P' process, which is also expected to be used to mass produced the A19 and A19 Pro next year for the iPhone 17 series. The chip itself could sport various chiplets, with each of them designed for a specific function. Apple could later combine each of these chiplets into a single unit, with Broadcom possibly included to help with how each of these processors communicates with each other when simultaneously running in Apple Intelligence servers. The reason why a chiplet design could be explored by the company is to reduce manufacturing complications, which would otherwise raise costs, but it could also allow Apple to keep the overall design hidden, even from its partners like Broadcom. As for the actual servers, we previously reported that Foxconn had been tasked to produce them, with Apple's assembling partner earlier said to receive some assistance from Lenovo and its subsidiary when it comes to design.

Apple reportedly working with Broadcom on custom AI chip, a boost for ASIC

According to the Information, Apple is reportedly developing a custom AI chip, codenamed Baltra, to power its AI services and devices. The tech giant is partnering with Broadcom to integrate advanced networking technology into the chip. This collaborative effort highlights Apple's commitment to in-house chip development and its increasing focus on AI. The chip is expected to be ready for mass production by 2026. In its efforts to catch up in the AI race, Apple has begun introducing its first generative AI features, collectively known as Apple Intelligence, for its iPhone, iPad, and Mac devices. While this is just the beginning, the company aims to expand its AI capabilities significantly. Apple is also investing in its own data centers equipped with custom-designed chips to power advanced AI tasks, ensuring a seamless AI experience across its product ecosystem. As part of its AI strategy, Apple is diversifying its chip reliance. At the latest AWS re:invent, Benoit Dupin, Apple's senior director of machine learning and AI, emphasized the decade-long partnership between Apple and AWS. During this collaboration, Apple has utilized various AWS technologies, including the Graviton3, Trainium2, and Inferentia2 chips, which have played a crucial role in enhancing Apple's machine-learning capabilities. These technologies support data processing, training, optimization, and the development of Apple Intelligence. Meanwhile, Bloomberg Intelligence notes that Apple's rumored collaboration with Broadcom further solidifies Broadcom's dominance in ASIC design. Broadcom already holds over 80% of the AI ASIC market and has a history of designing TPUs for Google. This partnership with Apple is expected to drive incremental growth in AI revenue for Broadcom beyond 2025-26 and likely expand its presence within Apple's supply chain. Adding to Broadcom's momentum, Reuters reported that OpenAI is also collaborating with Broadcom, a development that could further solidify Broadcom's dominant position in ASIC design.

Apple rumored to work with Broadcom on first server chip for AI, putting conversational Siri timeline in focus - 9to5Mac

In a heavily paywalled article, The Information reports that Apple is partnering with Broadcom to develop a new AI chip specifically designed for server use. Apple currently runs Apple Intelligence through its M-series chips on its servers through a system known as Private Cloud Compute. The distinction, according to the report, is that the new chip expected as soon as 2026, will be Apple's first server chip specifically designed for AI tasks. The Information cites "one of the people" in saying that Broadcom is Apple's partner for developing its server chip for artificial intelligence. The chip, The Information says, carries the code name Baltra. So why the Broadcom connection and why a new AI chip? Apple has designed its own processors for the iPhone and Mac for over a decade, and the Mac has ran on an Apple designed chip for years. These chips are designed with neural engines for processing AI tasks quickly. However, the rumored server chip for AI tasks can have specifications that don't make sense in a personal computer. Broadcom, a U.S. firm, specializes in manufacturing server processors. For Apple Intelligence, the rumor bodes well for the future of complex artificial intelligence tasks. Given the rumored 2026 target, this could mean Apple Intelligence could get much more capable sooner than later. An upgrade to Apple's Private Cloud Compute system could boost an already rumored Siri feature. What complexity could Apple Intelligence offer? Mark Gurman at Bloomberg reported last month that Apple intends to release a more conversational version of Siri that could be announced as soon as next year and launched as early as -- wait for it -- 2026. This version of Siri would offer the same type of conversational query input that ChatGPT and Claude provide today.

Apple is working on AI chip with Broadcom, the Information reports

Dec 11 (Reuters) - Apple (AAPL.O), opens new tab is developing its first server chip specially designed for artificial intelligence, the Information reported on Wednesday, citing three people with direct knowledge of the matter. The iPhone maker is working with Broadcom (AVGO.O), opens new tab on the chip's networking technology, which is crucial for AI processing, the report said. Apple and Broadcom did not immediately respond to Reuters' requests for comment. Following the report, shares of Broadcom were up 4.8% in premarket trading. Last year, Apple had signed a multi-billion-dollar deal with the chipmaker to develop 5G radio frequency components. Reporting by Harshita Mary Varghese in Bengaluru; Editing by Shilpi Majumdar Our Standards: The Thomson Reuters Trust Principles., opens new tab Suggested Topics:Technology

Apple Teams With Broadcom To Develop Its Own AI Chips and Launches Siri With ChatGPT

The iPhone maker is collaborating with chipmaker Broadcom (AVGO) to develop a chip code-named Baltra, Reuters reported Wednesday, citing coverage from The Information. Mass production of the chip is expected to begin in 2026, the report said. The chip will reportedly be made using an advanced manufacturing process from Taiwan Semiconductor Manufacturing Company (TSM), which produces chips for Nvidia and is an Apple supplier. Shares of Broadcom jumped over 6% in afternoon trading Wednesday, while Apple edged 0.2% higher, and TSMC shares gained 1.3%. Meanwhile, Apple Intelligence, the company's AI offering for iPhone 16 and other devices, got a series of updates Wednesday through iOS 18.2, including integration with OpenAI's generative AI chatbot, ChatGPT. Apple said the upgrade means its digital personal assistant, Siri, will be able to suggest users access ChatGPT in response to queries and also can provide the response itself, without switching apps. Customers can toggle on and off the ChatGPT integration, and OpenAI will not store inputs when used without a ChatGPT account, Apple said.

Apple Intelligence Servers Expected to Get All-New, Turbocharged Chip

Apple Intelligence servers are currently powered by the M2 Ultra chip, and they are expected to start using M4 series chips next year. In an eventual move away from Mac chips for server use, The Information today reported that Apple is developing a new server chip that will offer even faster performance for AI features. The report said the new chip contains "many duplicates" of Apple's Neural Engine, so it sounds like it will offer turbocharged performance for AI processing. At least a trio of companies are believed to be involved with the chip. Apple is said to be handling the overall design of the chip, while Broadcom is said to be providing some networking technology for it. TSMC is expected to begin mass production of the chip in 2026, using its third-generation 3nm process, known as N3P. That is the same process expected to be used for the A19 chips in the iPhone 17 models next year. The chip will likely power some Apple Intelligence features that rely on server-based generation, such as Image Playground, according to the report: Apple will likely use the new AI chip it is developing for inference, where the chip processes new data -- such as a user describing a possible image -- and applies them to the models to generate an output, like producing the image itself. More on Broadcom's involvement, from the report: Like Google, Apple is relying on Broadcom for technology to network or link the chips together so they can work in unison to compute data more quickly. That technology has been one of the key drivers of AI development, making it possible to compute the massive amounts of data required to train and run LLMs. Networking technology is one of Broadcom's key strengths. The paywalled report offers many more details about the chip. Apple has announced that iOS 18.2, iPadOS 18.2, and macOS Sequoia 15.2 will be released today with additional Apple Intelligence features, including Genmoji, Image Playground, Image Wand, and ChatGPT integration for Siri. There is also a new Visual Intelligence feature on all iPhone 16 models that allows you to quickly identify things in the real world using the Camera Control button on those devices. As part of Apple Intelligence, Siri will gain features like on-screen awareness and deeper per-app controls, likely starting with iOS 18.4. And with iOS 19.4 in 2026, Siri is expected to become more conversational like ChatGPT.