--- title: "Apple discusses hosting the new version of Siri by Google: Growing demand for cloud services" type: "News" locale: "en" url: "https://longbridge.com/en/news/277480839.md" description: "Apple is in discussions with Google to host the new version of Siri on Google Cloud to address its shortcomings in cloud computing capabilities. Apple has already utilized Google's Gemini large model in its products and hopes to enhance the performance of its AI assistant through Google's cloud services. Although Apple has attempted to establish its own cloud infrastructure, these efforts have repeatedly been hindered due to the attitudes of financial executives and insufficient investment. Apple's cloud computing strategy still relies on a combination of public cloud services and its own infrastructure" datetime: "2026-03-02T14:11:02.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/277480839.md) - [en](https://longbridge.com/en/news/277480839.md) - [zh-HK](https://longbridge.com/zh-HK/news/277480839.md) --- > Supported Languages: [简体中文](https://longbridge.com/zh-CN/news/277480839.md) | [繁體中文](https://longbridge.com/zh-HK/news/277480839.md) # Apple discusses hosting the new version of Siri by Google: Growing demand for cloud services In January of this year, Apple announced that it would use Google's Gemini large model in its products, which amounts to an acknowledgment that this iPhone manufacturer has been unable to compete in the field of artificial intelligence on its own in recent years. However, the agreement also indicates that Apple is once again relying on external companies to compensate for another shortcoming: cloud computing capabilities. Currently, Apple may further deepen its reliance on Google Cloud. According to two individuals familiar with the negotiations, at Apple's request, Google has explored deploying servers within its data centers to run the upcoming version of Siri. This Gemini-powered digital assistant will also comply with Apple's privacy standards. Apple is currently dependent on Google Cloud for services such as online storage and internal AI model training. In the past, Apple has stated that it would send complex AI query requests from users to a system called Private Cloud Compute, which runs on servers equipped with Apple's self-developed chips (simple AI queries are processed locally on the device). For years, Apple has been eager to achieve self-sufficiency in cloud computing. Several executives at the company have attempted to build internal infrastructure to reduce reliance on cloud service providers like Google and Amazon Web Services. However, according to a dozen former Apple executives and engineers, these cloud-related efforts have repeatedly faced obstacles because Apple's financial executives view cloud computing as a tedious cost center rather than a strategic priority. These individuals stated that Apple's reluctance to make significant investments in its own infrastructure has led to a continuous departure of cloud experts. In recent years, there has been increasing skepticism about Apple's insufficient investment in data centers compared to its peers, while Apple has assured investors that its "hybrid" infrastructure strategy—combining public cloud services with its own infrastructure—is performing well. Apple's predicament in cloud computing parallels its setbacks in the AI field and is interconnected in many ways. Over the past year, Apple has struggled to launch a comprehensive overhaul of Siri powered by AI on time, and the AI features that have been released have received lukewarm responses. Former employees revealed that the average utilization rate of Apple's private cloud computing is only 10%, with such low usage that some servers prepared for Apple's AI cloud are still stored in warehouses and have not been installed. However, if Apple claims that the new version of Siri launched this year is well-received by users, its demand for AI computing power may surge rapidly, which could explain why Apple is negotiating with Google about hosting the assistant. Apple's reliance on Google for cloud computing stands in stark contrast to the company's consistent philosophy of controlling core components of its products. Apple is known for its self-designed hardware, software, chips, and other key parts, which gives its devices an advantage over competitors At the same time, Apple is also known for its frugality, especially in projects like building data centers that require huge upfront capital expenditures. In recent years, companies like Meta, Microsoft, Google, and Amazon have made unprecedented investments in building data centers to accommodate the explosion of AI demand, while Apple has largely stayed out of it. In contrast, Apple's financial executives prefer to rent AI computing power and other services from external cloud service providers. Indeed, as demonstrated by OpenAI and Anthropic, owning data centers is not a prerequisite for becoming an AI leader. However, if cloud service providers choose to raise prices, excessive reliance on external companies may lead to regret, and AI companies may be forced to build more servers themselves. A former Apple cloud engineer stated that perhaps a more serious issue is that Apple's culture still revolves around device sales, which account for the vast majority of its revenue. Even if Apple earns more revenue from music, the App Store, and other services (which either run on its own cloud infrastructure or rent servers from other vendors), this has not changed. Apple's chaotic backend internal infrastructure further exacerbates the dilemma. Multiple departments within the company use their own servers or various cloud services, contrasting with the model of large tech companies like Google, which unify computing resource pools for engineers to access. Igor Nafnyuk, who participated in the development of the next-generation Siri and left last December, previously worked in Google's infrastructure department for over a decade. He stated, "Apple's engineering culture is completely different from Google's. Most of Google's systems are centralized, with everyone using the same supercomputer; whereas at Apple, technology choices are fragmented and operate independently." **Strict Cost Control** Apple's cloud dilemma can be traced back several decades. After the launch of the iTunes Music Store in 2003, Apple began to expand the scale of small data centers to support the burgeoning new business of digital music sales. As Apple's online services continued to expand, it used a variety of systems to operate separately. For example, the personalized playlist feature iTunes Genius, launched in 2008, was built on technology and servers independent of iTunes because Apple wanted to anonymize the analysis of user music library data. Around the same time, Apple began to increasingly rely on public cloud services—a novel concept at the time pioneered by Amazon Web Services (AWS), which rents large data centers to external enterprise customers. AWS was an early online storage provider for Apple's iCloud storage service launched in 2011. However, when Apple planned to launch the iCloud photo backup feature, the high costs of public cloud storage became overwhelming as a large number of iPhone users began taking photos. To save costs, Apple started building its own servers for iCloud photo backup. This move also brought another benefit: according to a former Apple executive involved in the project, when Apple informed AWS of its plans, Amazon halved the cloud service fees it would charge Apple. Apple then continued to use both its own servers and those of cloud service providers By 2013, Apple's finance department was dissatisfied with the skyrocketing costs associated with public cloud and its own servers, and began to inquire whether service teams were fully utilizing the company's existing internal infrastructure. Former Apple engineers stated that these teams clearly lacked coordination in server deployment, leading to redundant infrastructure and idle resources. For example, when the iTunes servers had idle cloud capacity, other Apple teams were unable to use it. In 2013, Apple appointed then-Engineering Director Patrick Gates to integrate the decentralized server infrastructure into a public resource pool accessible to all teams. He led the establishment of a new department called "Platform Infrastructure Engineering," building a shared resource model based on the modern cloud systems of Amazon and Google. **ACDC Project** However, according to former Apple engineers, Gates struggled to promote the use of this centralized platform across the company's departments and ultimately left in 2018. In 2019, Mike Abbot, former Vice President of Engineering at Twitter and early engineering lead at Microsoft Cloud, took over the department and continued to advance the project. Abbot attempted to cultivate a more cloud-service-oriented culture within Apple. In 2021, he initiated an internal "Infrastructure Summit" online event aimed at promoting collaboration across the company on shared infrastructure. He also pushed for several new initiatives, the most notable being the ACDC project, which aimed to apply Apple chip technology to its own data center servers. The full name of the project is "Apple chips in data center," with the intention of ensuring that data centers running Apple’s online services meet the same strict privacy standards as Apple devices. According to previous reports, he also suggested that Apple consider eventually renting servers to external developers, similar to Amazon and Google's public cloud services. The former project lead stated that Apple chip chief Johny Srouji was a strong supporter of this project, believing it was an opportunity to provide enterprise customers with chips developed by his team. However, former team employees indicated that several of Abbot's projects faced difficulties due to opposition from the finance department. The finance team believed that the existing server utilization was insufficient and was unwilling to invest further in Apple's own cloud services, believing that relying on external cloud service providers could better control infrastructure costs. In 2023, Abbot left Apple to join General Motors. His departure quickly triggered a talent exodus in Apple's cloud sector, with many of the employees he recruited following him to the automaker. **Turning to Google Cloud** The launch of ChatGPT 3.5 at the end of 2022 was a disruptive moment for the entire tech industry and fundamentally changed Apple's approach to AI and cloud. Before the release of ChatGPT, Apple's core goal for Siri was to operate as efficiently as possible, prioritizing local processing on devices rather than in the cloud, as Apple believed this would better protect user data privacy. After ChatGPT demonstrated the immense potential of cloud-based large models, Apple realized it could no longer cling to its old model and needed to leverage cloud computing power more But the problem arose: Apple's internal AI infrastructure began to age. According to relevant sources, in 2023, the company started to phase out a large number of aging NVIDIA chips in its data centers that were gradually failing, a task that had been planned for years but had been postponed. Apple needs to replace these old chips with newer models that are more compatible with the latest AI technologies. Driven by the finance department, Apple chose to primarily rely on external cloud service providers to advance its AI plans, similar to its approach in storage and other services. Long-term cloud service provider AWS became Apple's first important partner in its AI layout, and Apple was also an early customer of Amazon's self-developed NVIDIA AI chip alternatives, Inferentia and Trainium2. The collaboration with Google was more complicated. For many years, due to privacy concerns, Apple prohibited AI engineers from using Google Cloud. Siri handles personal identity information of Apple device users, and Apple was unwilling to take on any risk of data leakage to external companies. In fact, Craig Federighi, the software executive who serves as Apple's privacy chief, repeatedly vetoed proposals to use Google Cloud to meet AI computing needs. However, in 2023, Google upgraded its security systems to meet Apple's privacy requirements. Apple quickly began to adopt Google Cloud to meet its AI needs, including using Google's custom chip tensor processing units—Apple calculated that their operating costs were far lower than those of comparable NVIDIA chips. **Private Cloud** Apple needs to prove to the outside world that it is taking AI seriously. The explosive growth of ChatGPT forced Google, Amazon, and almost all major tech companies to launch smarter and more conversational AI products. At the 2024 Apple Worldwide Developers Conference, Apple finally took action, announcing the launch of Apple Intelligence—a suite of AI tools based on the generative models popularized by ChatGPT. Apple assigned the ACDC project team, previously led by Abbott, to assist in the launch of Apple Intelligence. Although the project was not initially focused on AI, the team urgently built a private cloud computing system codenamed "Thimble Project" to provide more private computing support for Apple's next-generation AI products. Former employees involved in the project stated that although Apple announced the private cloud computing system alongside Apple Intelligence in June 2024, the system was not actually operational at that time, lagging six months behind schedule (it ultimately went live at the end of 2024). In the following months, Apple gradually rolled out some features of Apple Intelligence, such as AI writing tools and notification summaries, but the public and tech commentators generally felt disappointed. For Apple, a more serious issue was that a revamped, more conversational version of Siri was still unable to be launched. The negotiations between Apple and Google regarding the hosting of Siri may indicate that Apple hopes to be prepared for a surge in device-side AI activities after the new version of Siri is launched later this year Another reason behind the discussions may be that private cloud computing is not performing well within Apple's own data centers. Former employees stated that software updates for its AI servers take much longer than for other types of servers. Additionally, Apple cloud engineers and former employees from the AI department indicated that the Apple chips on the private cloud computing servers are not designed for AI, making it difficult to run large models like Google Gemini effectively ### Related Stocks - [Apple Inc. (AAPL.US)](https://longbridge.com/en/quote/AAPL.US.md) - [Alphabet Inc. (GOOG.US)](https://longbridge.com/en/quote/GOOG.US.md) - [Alphabet Inc. (GOOGL.US)](https://longbridge.com/en/quote/GOOGL.US.md) - [Direxion Daily GOOGL Bull 2X Shares (GGLL.US)](https://longbridge.com/en/quote/GGLL.US.md) - [GraniteShares 2x Long AAPL Daily ETF (AAPB.US)](https://longbridge.com/en/quote/AAPB.US.md) - [Direxion Daily GOOGL Bear 1X ETF (GGLS.US)](https://longbridge.com/en/quote/GGLS.US.md) - [T-Rex 2X Long Apple Daily Target ETF (AAPX.US)](https://longbridge.com/en/quote/AAPX.US.md) - [Roundhill GOOGL WeeklyPay ETF (GOOW.US)](https://longbridge.com/en/quote/GOOW.US.md) - [Global X Cloud Computing ETF (CLOU.US)](https://longbridge.com/en/quote/CLOU.US.md) - [Direxion Daily AAPL Bear 1X Shares (AAPD.US)](https://longbridge.com/en/quote/AAPD.US.md) - [Direxion Daily AAPL Bull 2X Shares (AAPU.US)](https://longbridge.com/en/quote/AAPU.US.md) - [Kurv Yield Premium Str Apple ETF (AAPY.US)](https://longbridge.com/en/quote/AAPY.US.md) - [YieldMax AAPL Option Income Strategy ETF (APLY.US)](https://longbridge.com/en/quote/APLY.US.md) ## Related News & Research - [Apple Inc. $AAPL is Steadtrust LLC's 3rd Largest Position](https://longbridge.com/en/news/280017068.md) - [Cambria Investment Management L.P. Has $4.55 Million Position in Alphabet Inc. $GOOGL](https://longbridge.com/en/news/279906352.md) - [Google Bets On AI To Simplify Prescriptions And Symptom Checks In Europe](https://longbridge.com/en/news/279794182.md) - [Google to allow AI opt-out to ease UK competition concerns](https://longbridge.com/en/news/279630314.md) - [Alphabet Inc. $GOOGL Shares Purchased by Park Avenue Securities LLC](https://longbridge.com/en/news/280046998.md)