--- title: "Locking in Apple ASIC orders until 2031! Broadcom leaps from RF chip supplier to \"AI factory\" level full-stack seller" type: "News" locale: "en" url: "https://longbridge.com/en/news/291854364.md" description: "Broadcom and Apple expand cooperation, providing Apple with new custom high-performance ASIC chips, with the collaboration extended until 2031. The agreement involves multiple generations of Apple products and aims to support artificial intelligence training and inference tasks. As a core chip supplier, Broadcom's AI ASIC business has become a key pillar of valuation growth, and recent financial reports have confirmed this strong growth logic" datetime: "2026-07-07T00:32:04.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/291854364.md) - [en](https://longbridge.com/en/news/291854364.md) - [zh-HK](https://longbridge.com/zh-HK/news/291854364.md) --- # Locking in Apple ASIC orders until 2031! Broadcom leaps from RF chip supplier to "AI factory" level full-stack seller According to the Zhitong Finance APP, Broadcom (AVGO.US), one of the big winners in the global AI boom, announced that it will provide Apple Inc. (AAPL.US) with new customized high-performance chips as part of an expanded partnership that will last until 2031. Under the newly signed agreement, the two tech giants will engage in deep cooperation on ASIC chips (Application-Specific Integrated Circuits), which are specialized integrated circuits. According to a document submitted by Broadcom on Monday local time, these ASIC chips will be used in "multiple generations of Apple products." ASIC chips are silicon chips developed for specific purposes, and they are becoming increasingly important for developing AI chip components and clusters used for processing AI-related training/inference tasks. Media reports have recently indicated that Apple is developing more advanced AI chips or complete AI servers for AI training/inference, with plans to deploy these AI server superclusters as early as 2027. Broadcom is one of the core chip suppliers for Apple and other large tech companies, as well as a key supplier of high-performance Ethernet switch chips and high-speed optical interconnect core devices for large AI data centers, along with customized AI chips that are crucial for AI training/inference for cloud computing giants. Broadcom has tapped into the massive incremental market for customized AI chips beyond NVIDIA GPUs, focusing on cloud giants like Google TPU (a typical AI ASIC technology route) and OpenAI/Anthropic's self-developed AI ASIC accelerators. AI ASICs are undoubtedly the first core pillar for Broadcom's valuation and fundamental growth. The strong financial reports recently released by Broadcom and Marvell demonstrate that the unprecedented strong growth logic of AI ASICs is being rapidly confirmed by "earnings report-level evidence." Broadcom expects AI chip sales to exceed $100 billion by 2027, with Q2 FY2026 AI semiconductor revenue expected to reach $10.8 billion, a year-on-year increase of 143%, indicating that the demand for AI ASIC/XPU has become its core growth curve. On the other hand, all GPU, TPU, and ASIC routes rely on high-speed Ethernet switching and optical interconnects. Broadcom's Tomahawk 6 has achieved a single-chip switching capacity of 102.4 Tbps and showcased 102.4T Ethernet, CPO, 400G/lane optical DSP core devices, PCIe Gen6, and other AI interconnect solutions at OFC 2026, precisely addressing the bottleneck of transitioning AI data centers from "single-chip computing power" to "ten-thousand-card/one-hundred-thousand-card cluster efficiency." Therefore, Broadcom's new round of strong bull market logic is not simply about "challenging NVIDIA," but rather about becoming a supplier of massive customized cloud AI inference workload computing resources, data center optical interconnect systems, and network infrastructure bases at the "AI factory" level: benefiting from collaborating with Apple to create ASIC components and working with cloud computing companies to significantly reduce costs for self-developed AI ASIC chips, while also benefiting from the strong demand for optical interconnects, Ethernet switching, and data center network infrastructure driven by all types of AI computing clusters Wall Street analysts have recently raised Broadcom's target stock price, primarily supported by the rapid growth of revenue related to AI ASIC and AI Ethernet switching infrastructure; VMware software cash flow, Ethernet switching, optical interconnect/optical communication, and data center interconnect form the second layer of support for Broadcom's AI infrastructure platform revaluation. ## Broadcom and Apple to collaborate until 2031, custom chips at the forefront of the tech giants' computing arms race! Broadcom's stock price surged over 6% at one point. For a long time, Broadcom has partnered with Apple to jointly design and manufacture core components related to radio frequency chips, cellular connectivity, Wi-Fi, and Bluetooth. However, in recent years, as Apple developed its own N1 chip, the scale of this collaboration has significantly reduced. Apple's exclusively self-developed N1 chip is a core hardware component that integrates Wi-Fi and Bluetooth, found in the recently mass-produced iPhone, iPad, and Mac. After the New York stock market opened, Broadcom's stock price rose by 6.3% to $383.16, ultimately closing up 3.73% at $373.90. Driven by the surge in chip demand due to the AI boom, the stock has increased by about 38% over the past year. The company has been deeply collaborating with several large tech companies, including Google, a subsidiary of Alphabet Inc, and Meta Platforms Inc., the parent company of Facebook. BNP Paribas recently raised its target price for Broadcom from $600 to $640 while maintaining an "outperform" rating, corresponding to a potential upside of about 71.2% and a potential market value of approximately $3.03 trillion. ![image.png](https://imageproxy.pbkrs.com/https://img.zhitongcaijing.com/image/20260707/1783381968814397.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg) The business expansion of major clients such as OpenAI, Meta, and Anthropic is transforming Broadcom from a "Google TPU supply chain + AI ASIC leader" into a "custom AI chips + network infrastructure + data center optical interconnect systems" provider, along with VMware's super platform-based AI computing infrastructure company. This is also the core logic behind Wall Street's continued push to raise Broadcom's target stock price. It is understood that Broadcom is still actively supplying some components related to cellular networks, which can be used with Apple's self-developed modem; the last time the two companies announced a collaboration project was in 2023. Senior analysts indicate that Apple accounts for about 20% of Broadcom's annual revenue. Emarketer analyst Jacob Byrne stated, "For Apple, signing a contract with Broadcom before 2031 can ensure supply chain stability amid the ongoing shortage of AI chips and avoid Apple having to procure the most critical components for iPhones on a large scale." "For Broadcom, this provides reassurance, as Apple has been trying to manufacture these chips on its own for years." Jacob Byrne added that this alleviates market concerns about Apple replacing Broadcom with its own solutions, and it shows that even as Apple continues to advance its self-developed modems, M-series processors, and AI server chips, its complex customized ASIC silicon, as well as its supply capabilities for RF, networking, and system-level chips, still require collaboration with top external suppliers. Apple has long focused on developing its first AI server chip specifically designed for massive AI workloads, internally codenamed "Baltra." These chips are a significant upgrade over the M5 Ultra chip, which Apple previously planned to launch in the Mac Studio product line this year. Apple is currently still using standard Mac chips in its AI server cluster focused on Apple Intelligence-related AI workloads. The upcoming AI server aims to provide more powerful processing capabilities and higher, more stable reliability for the cloud system components of Apple's exclusive AI application ecosystem platform—Apple Intelligence. Apple has been rapidly expanding its AI application capabilities, including the announcement of a new version of AI Siri that will embed AI voice features, utilizing a more powerful cloud AI large model designed by Google. ## A second AI computing power supply curve emerges beyond NVIDIA GPUs! AI computing infrastructure enters the "custom silicon era" Broadcom and Apple have extended their collaboration until 2031, which on the surface appears to be a supply chain renewal, but in essence marks Broadcom's positioning in the AI era as having upgraded from a "connectivity chip supplier" to a super-platform AI computing infrastructure company providing "AI AISC customized AI chips + network infrastructure + data center optical interconnect systems." The key for Apple is not simply to "move away from NVIDIA," but to build a more controllable, lower-cost, and better-adapted computing foundation for Apple Intelligence, cloud Siri, and future Baltra-like AI servers; Broadcom, on the other hand, connects the growth curves of consumer electronics terminals, edge AI, and cloud AI servers through Apple orders. The larger investment mainline is in cutting-edge AI clients such as Google, OpenAI, Anthropic, and Meta. Wall Street financial giant JP Morgan refuted rumors of delays for TPU v9, stating that the TPU v9 project managed by Broadcom remains one of the highest priority projects, and Google's internal chip team is still over 18 months behind Broadcom in advanced chip and packaging capabilities; more critically, the five-year agreement signed between Google and Broadcom in March covers TPU v8 to v11 and extends the revenue visibility related to Broadcom's TPU until 2031. JP Morgan expects Broadcom's AI-related revenue to grow 2 to 2.5 times by 2027 and to double again in 2028; this means that the market should not only view Broadcom as a "replacement transaction for NVIDIA's AI computing cluster," but rather as a core platform AI computing infrastructure supplier for AI ASIC outsourcing design, advanced packaging, SerDes, high-speed interconnects, and Ethernet cluster networks The AI computing power supply transaction between Broadcom and Anthropic indicates that AI computing power construction has transitioned from simple chip procurement to a financial engineering stage involving "chips + networks + private credit + long-term leases." Broadcom, Apollo, and Blackstone have established the AI XPV platform, with the first transaction of $35 billion aimed at supporting Anthropic's expansion of over 1 gigawatt of computing power infrastructure. The platform's goal is to support over 20 gigawatts of computing capacity for cutting-edge AI laboratories through Broadcom's XPU and network solutions by 2028. The investment implications of this structure highlight that AI chips are no longer just a revenue item for semiconductor companies but are being packaged as infrastructure assets with long-term contract cash flows, financing potential, and securitization; conversely, it is also clear that if AI-related revenue or profits materialize slower than computing power capital expenditures, supplier guarantees, SPV debt, customer credit, and chip residual value will all become risk points for the market to reassess valuations. OpenAI and Broadcom's launch of Jalapeño, an AI ASIC computing power infrastructure system, brings the logic of "custom/self-developed AI ASICs replacing part of the general GPU inference workload" to the forefront. OpenAI officially states that Jalapeño is its first self-developed intelligent processor aimed at large language model inference, taking only nine months from design to tape-out. Early tests show that its performance per watt significantly exceeds the current state-of-the-art level, and it plans to conduct gigawatt-level deployments with data center partners like Microsoft starting in 2026; this is not a comprehensive challenge to NVIDIA on the training side but rather a soft and hardware collaborative optimization around ChatGPT, Codex, API, and future intelligent agent products on the inference side, aiming to reduce the delivery cost per token and enhance throughput and reliability. The core significance of the latest launch of Jalapeño by OpenAI and Broadcom is not simply to "bypass NVIDIA," but rather that leading AI large model companies are beginning to extend computing power sovereignty from GPU procurement to a full-stack collaboration of "model - compiler - chip - network - data center optical interconnect." Jalapeño is positioned as OpenAI's first "Intelligence Processor" specifically for large language model inference, with engineering samples already running machine learning workloads including GPT-5.3-Codex-Spark at target frequencies and power consumption within just nine months from design to tape-out. OpenAI and Broadcom claim that its early performance per watt significantly exceeds the current state-of-the-art level and plan to enter gigawatt-level deployments with data center partners like Microsoft starting at the end of 2026. An earlier 10GW cooperation framework also shows that OpenAI designs accelerators and systems, while Broadcom is responsible for implementation, Ethernet infrastructure expansion, and high-speed optical connection solutions within data centers, with deployment starting in the second half of 2026 and continuing until the end of 2029. ![image.png](https://imageproxy.pbkrs.com/https://img.zhitongcaijing.com/image/20260707/1783382061936939.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg) Wall Street investment institutions, including BNP Paribas, unanimously believe that Broadcom is becoming the most critical custom ASIC silicon contractor for AI computing power outside of NVIDIA. JP Morgan raised its target price from $500 to $580, BofA's target price is $530, Evercore ISI's target price is $582, and HSBC has even raised its target price to $600. TIPRANKS shows that the average target price on Wall Street is as high as $516.91 (implying a potential increase of 38%). Analysts' core bullish points all point to the growth in ASIC revenue and the large-scale volume from multiple major clients in Broadcom's exclusive VMware virtual cloud, Ethernet infrastructure, data center optical interconnect/optical communication, and the interconnection ecosystem between multiple data centers (DCI). Tristan Gerra, a senior analyst at Robert W. Baird, who gives an optimistic target price of $630, stated that his core bullish logic focuses on three points: First, Broadcom's AI-related revenue guidance for 2026-2027 is stronger than market expectations and driven by multiple AI clients' volume; second, Broadcom has platform-level advantages in custom AI ASICs, Ethernet switch chips, and data center optical interconnect products; third, VMware's software assets provide cash flow and margin support, making Broadcom not just an "AI chip resilient stock," but a composite AI infrastructure company of "AI ASIC + Ethernet switch chips + data center optical interconnect + enterprise software cash flow." ### Related Stocks - [AVGO.US](https://longbridge.com/en/quote/AVGO.US.md) - [AAPL.US](https://longbridge.com/en/quote/AAPL.US.md) - [AVGU.US](https://longbridge.com/en/quote/AVGU.US.md) - [AAPB.US](https://longbridge.com/en/quote/AAPB.US.md) - [AVGX.US](https://longbridge.com/en/quote/AVGX.US.md) - [AVGW.US](https://longbridge.com/en/quote/AVGW.US.md) - [AVS.US](https://longbridge.com/en/quote/AVS.US.md) - [AVL.US](https://longbridge.com/en/quote/AVL.US.md) - [AVGG.US](https://longbridge.com/en/quote/AVGG.US.md) - [AAPW.US](https://longbridge.com/en/quote/AAPW.US.md) - [SMH.US](https://longbridge.com/en/quote/SMH.US.md) - [AAPD.US](https://longbridge.com/en/quote/AAPD.US.md) - [AAPU.US](https://longbridge.com/en/quote/AAPU.US.md) - [AAPY.US](https://longbridge.com/en/quote/AAPY.US.md) - [AAPX.US](https://longbridge.com/en/quote/AAPX.US.md) - [MAGX.US](https://longbridge.com/en/quote/MAGX.US.md) - [APLY.US](https://longbridge.com/en/quote/APLY.US.md) - [SOXX.US](https://longbridge.com/en/quote/SOXX.US.md) - [SOXL.US](https://longbridge.com/en/quote/SOXL.US.md) - [SOXQ.US](https://longbridge.com/en/quote/SOXQ.US.md) - [XSD.US](https://longbridge.com/en/quote/XSD.US.md) - [PSI.US](https://longbridge.com/en/quote/PSI.US.md) - [FTXL.US](https://longbridge.com/en/quote/FTXL.US.md) - [MRVL.US](https://longbridge.com/en/quote/MRVL.US.md) - [NVDA.US](https://longbridge.com/en/quote/NVDA.US.md) - [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md) - [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md) - [META.US](https://longbridge.com/en/quote/META.US.md) - [BNPQY.US](https://longbridge.com/en/quote/BNPQY.US.md) - [JPM.US](https://longbridge.com/en/quote/JPM.US.md) - [BAC.US](https://longbridge.com/en/quote/BAC.US.md) - [EVR.US](https://longbridge.com/en/quote/EVR.US.md) - [HSBC.US](https://longbridge.com/en/quote/HSBC.US.md) - [00005.HK](https://longbridge.com/en/quote/00005.HK.md) - [APO.US](https://longbridge.com/en/quote/APO.US.md) - [BX.US](https://longbridge.com/en/quote/BX.US.md) - [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md) - [NVD.DE](https://longbridge.com/en/quote/NVD.DE.md) - [JPM-M.US](https://longbridge.com/en/quote/JPM-M.US.md) - [JPM-C.US](https://longbridge.com/en/quote/JPM-C.US.md) - [JPM-D.US](https://longbridge.com/en/quote/JPM-D.US.md) - [JPM-L.US](https://longbridge.com/en/quote/JPM-L.US.md) - [8634.JP](https://longbridge.com/en/quote/8634.JP.md) - [JPM-K.US](https://longbridge.com/en/quote/JPM-K.US.md) - [JPM-J.US](https://longbridge.com/en/quote/JPM-J.US.md) - [BAC-Q.US](https://longbridge.com/en/quote/BAC-Q.US.md) - [BML-L.US](https://longbridge.com/en/quote/BML-L.US.md) - [BAC-L.US](https://longbridge.com/en/quote/BAC-L.US.md) - [BAC-K.US](https://longbridge.com/en/quote/BAC-K.US.md) - [BML-H.US](https://longbridge.com/en/quote/BML-H.US.md) - [BAC-N.US](https://longbridge.com/en/quote/BAC-N.US.md) - [BAC-E.US](https://longbridge.com/en/quote/BAC-E.US.md) - [MER-K.US](https://longbridge.com/en/quote/MER-K.US.md) - [BAC-O.US](https://longbridge.com/en/quote/BAC-O.US.md) - [BML-G.US](https://longbridge.com/en/quote/BML-G.US.md) - [BAC-M.US](https://longbridge.com/en/quote/BAC-M.US.md) - [BAC-B.US](https://longbridge.com/en/quote/BAC-B.US.md) - [BML-J.US](https://longbridge.com/en/quote/BML-J.US.md) - [BAC-S.US](https://longbridge.com/en/quote/BAC-S.US.md) - [BAC-P.US](https://longbridge.com/en/quote/BAC-P.US.md) - [8648.JP](https://longbridge.com/en/quote/8648.JP.md) - [DTIW.SG](https://longbridge.com/en/quote/DTIW.SG.md) - [HSBA.UK](https://longbridge.com/en/quote/HSBA.UK.md) - [APO-A.US](https://longbridge.com/en/quote/APO-A.US.md) - [APOS.US](https://longbridge.com/en/quote/APOS.US.md) ## Related News & Research - [Broadcom’s Largest AI Customer Is Fleeing to MediaTek. AVGO Stock Is Still a Buy.](https://longbridge.com/en/news/291836733.md) - [Broadcom Earns Buy Rating on Expanding $30B+ Apple Partnership, AI-Driven Growth, and Attractive Valuation](https://longbridge.com/en/news/292108066.md) - [Meta's $6.5B AI Chip Shift Could Boost Samsung](https://longbridge.com/en/news/291931335.md) - [Tesla, Meta to Anthropic: Samsung's chip orders reach 50 trillion won.](https://longbridge.com/en/news/291693528.md) - [Broadcom, Apple extend chip partnership through 2031](https://longbridge.com/en/news/291801241.md)