Google Shifts Gears: Potential Partnership with MediaTek for Next-Gen AI Chips

Google preparing to partner with Taiwan's MediaTek on next AI chip, Information reports

March 17 (Reuters) - Alphabet's (GOOGL.O), opens new tab Google is preparing to partner with Taiwan's MediaTek (2454.TW), opens new tab on the next version of its AI chips, Tensor Processing Units, that will be made next year, the Information reported on Monday, citing people involved in the project. However, Google has not cut ties with Broadcom (AVGO.O), opens new tab, the chip designer it has worked with exclusively on the AI chips over the past several years, the report said, citing an employee at the San Jose-based company. Google, MediaTek and Broadcom did not immediately respond to Reuters' requests for comment. Like Nvidia (NVDA.O), opens new tab,. Google also designs its own AI server chips, which it uses for internal research and development and also rents out to cloud customers. This approach gives Google a competitive edge in the AI race by reducing its reliance on Nvidia, even as rivals like Microsoft-backed (MSFT.O), opens new tab OpenAI and Meta Platforms (META.O), opens new tab have seen a surge in demand for Nvidia chips. Late last year, Google rolled out its sixth-generation TPU in a bid to give itself and its Cloud customers an alternative to Nvidia's chips, which are the most sought-after processors in the industry. Google chose MediaTek partly because the Taiwanese firm has a strong relationship with TSMC (2330.TW), opens new tab and charges Google less per chip compared to Broadcom, the Information report added. Google spent between $6 billion and $9 billion on TPUs last year, according to research firm Omdia, based on Broadcom's target for AI semiconductor revenue last year. Reporting by Akash Sriram and Deborah Sophia in Bengaluru; Editing by Maju Samuel Our Standards: The Thomson Reuters Trust Principles., opens new tab Suggested Topics:Artificial Intelligence

Google could partner with MediaTek for next gen AI chips: Report

Google has filled Android with 'recommendations' to the point they are now unbearable Summary Google may switch TPU partners from Broadcom to MediaTek for cost optimization. MediaTek might handle I/O modules in partnership, with Google focusing on more in-house TPU design. Changing partners could reduce Google's Broadcom dependency in the AI market. AI models are computation-intensive, requiring powerful compute units which Google calls Tensor Processing Units (TPUs). Its Gemini chatbot and other AI tools are available to most users for free, and only a small subset of AI features are paywalled, limiting access to Workspace customers. So, it is understandable if the company wishes to optimize the costs associated with acquiring the processing horsepower. With AI and demand for the tech showing no signs of letting up, a new report based on insider reports states that Google may consider a new TPU production partner in chipmaker MediaTek, after favoring Broadcom for the past nine years. Related What are Google Cloud TPUs? Google's aiming for the cloud-based AI creation sector: Google Cloud TPUs are helping it win Posts Since 2015, Google has used Broadcom TPUs for AI. While the first generation hardware has been retired, TPUs from 2017 through 2021 are currently serving cloud customers. Subsequent generations of the chip have enabled Google engineering efforts while also serving cloud customers. 2025's will soon be available and according to a report from The Information, it could be the last generation exclusively handled by Broadcom. The tech partner co-designs Google's AI chips based on blueprints the latter supplies, and Taiwanese foundry TSMC handles production, a Broadcom staffer reportedly said. An unnamed source at TSMC added that MediaTek charges Google less per chip and also has strong ties with the same foundry. Retaining a cost advantage would help Google stay competitive in a market where rivals like OpenAI have been buying NVIDIA's chips to power AI applications like ChatGPT. That doesn't mean Google isn't an NVIDIA customer. It placed an order for over $10 billion worth of Blackwell chips. Separately, the company spent an estimated $6-9 billion on TPUs last year alone. MediaTek will only handle I/O modules Nothing set in stone yet Source: Google Blog An important detail outlined in the report is that MediaTek will only handle the design of the I/O module designs in the planned partnership, while Google will design most of the TPU in-house. The new partner will also liaise with TSMC and be responsible for quality control. This is a shift for the Search titan that had depended on Broadcom until now for all the heavy lifting. Moving more operations in-house while using the same manufacturer in TSMC could prove beneficial for Google in the long term. Google seems to be doubling down on in-house TPU design since it is recruiting chip design engineers in Taiwan, including former TSMC staffers. However, the company reportedly produces two different TPUs -- one for training new models and another to power services like Search, YouTube, and Gemini. It's still unclear if the company's internal strategy will shift to making a single chip if MediaTek is onboarded as a partner for the 2026 production. While the impact on consumer-facing AI applications will be none to minimal, we suspect Google could reap the benefits of reducing its Broadcom dependency in the coming years as demand for AI compute modules continues to soar. That said, Broadcom has other big-ticket clients in Meta Platforms, Arm Holdings, Apple, and OpenAI, so it isn't going under anytime, even though Google is the company's biggest revenue source from the sale of AI chips. Embed Gemini here

Google reportedly partnering with MediaTek for next-generation TPU production - SiliconANGLE

Google reportedly partnering with MediaTek for next-generation TPU production Google LLC is reportedly preparing to partner with Taiwanese fabless semiconductor company MediaTek Inc. on the next version of its Tensor Processing Units, the company's in-house artificial intelligence chips. According to The Information referencing an internal source, the next-generation TPUs could be produced by MediaTek starting sometime next year. The report also claims that processing could be done differently, with MediaTek handling input/out modules that manage the communication between the main processors and peripheral components. Google currently uses Broadcom Inc. to produce its TPU chips and while MediaTek may start producing the next generation of chips, it's noted that Google hasn't cut ties with Broadcom, at least for the time being. Google's TPUs are custom-designed application-specific integrated circuits developed to accelerate machine learning tasks, particularly those involving neural networks. The chips are optimized for Google's TensorFlow framework and enhance both training and inference processes by efficiently handling the computational demands of deep learning models. TPUs differ from traditional processors by being tailored for high-volume, low-precision arithmetic operations. The specialization enables TPUs to deliver significant performance improvements and energy efficiency compared to general-purpose CPUs and GPUs. In cloud environments, TPUs are integrated into Google's data centers to provide scalable and efficient resources for large-scale machine-learning tasks. For edge computing, Google offers the Edge TPU, a compact and power-efficient version designed to bring AI capabilities to devices like smartphones and IoT applications. While TPUs have various purposes, AI is among their best-known applications. The specialized processors are designed to accelerate machine learning tasks, particularly those involving neural networks, by efficiently handling the massive computations required for training and inference. TPUs enable faster processing and improved energy efficiency by offloading intensive workloads from general-purpose CPUs and GPUs, resulting in the development and deployment of complex AI models across various industries. As noted by Reuters, being able to produce in-house AI chips also gives Google a competitive edge in the AI race by reducing its reliance on Nvidia," with Nvidia Corp. currently dominating the market for AI chips. The most recent release of Google TPUs was the sixth generation, the Trillium TPU, which was announced in October. Offering an alternative to Nvidia's popular GPUs, Trillian provides a four-times performance boost in terms of AI training and a three-times boost in inference throughput compared to the previous generation of chips. Trillian TPUs also include increased memory and bandwidth allowing the chips to run much larger large language models with more weights and larger key-value caches. The chip supports a broader range of model architectures in both training and inference.

Google to Partner With MediaTek for Next-Gen TPU

Google is set to partner with Taiwanese chip maker MediaTek for the next version of its Tensor Processing Units (TPUs), reported The Information on Monday. This marks a significant shift as Google has relied on Broadcom as the primary supplier of TPUs for nearly a decade. However, it was also reported that Google has not cut ties with Broadcom despite the new development. Google's decision comes against the backdrop of MediaTek's ties with TSMC. The AI giant is looking to have cost leverage over chip production, the report said. Furthermore, Google is likely to handle the design of the TPU for the most part, while MediaTek will handle the input and output modules. This approach differs from Google's collaboration with Broadcom, which developed the core parts of the TPU chips, according to sources cited by The Information. Google's TPUs are custom-designed chips built to accelerate machine learning workloads across various platforms, from large-scale data centres and individual edge devices alongside cloud-based systems. TPUs power Google's leading AI models, applications and services. In 2023, The Information also reported that Google's executives 'extensively discussed dropping Broadcomm' as the supplier of TPUs as early as 2027. According to a report from Omdia last year, Google spent $6 billion to $9 billion on TPUs. The company announced its sixth generation of the TPU - Trillium in 2024. Compared to its predecessor, Trillium delivers a four-fold performance improvement. Further, it also provides a 67% increase in energy efficiency and a three-fold increase in inference throughput. In January, an industry analyst pegged the value of a 'potential stand-alone TPU business' to be over $700 billion. Google's TPUs have been an alternative to several companies over the chips sold by NVIDIA. Last year, it was also reported that Apple employed Google's TPU clusters to train its foundational models.

Google Preparing to Partner with Taiwan's MediaTek on Next AI Chip: Report

Google has not cut ties with Broadcom for AI chips Like Nvidia,. Google also designs its own AI server chips Google rolled out its sixth-generation TPU late last year Alphabet's Google is preparing to partner with Taiwan's MediaTek on the next version of its AI chips, Tensor Processing Units, that will be made next year, the Information reported on Monday, citing people involved in the project. However, Google has not cut ties with Broadcom, the chip designer it has worked with exclusively on the AI chips over the past several years, the report said, citing an employee at the San Jose-based company. Google, MediaTek and Broadcom did not immediately respond to Reuters' requests for comment. Like Nvidia, Google also designs its own AI server chips, which it uses for internal research and development and also rents out to cloud customers. This approach gives Google a competitive edge in the AI race by reducing its reliance on Nvidia, even as rivals like Microsoft-backed OpenAI and Meta Platforms have seen a surge in demand for Nvidia chips. Late last year, Google rolled out its sixth-generation TPU in a bid to give itself and its Cloud customers an alternative to Nvidia's chips, which are the most sought-after processors in the industry. Google chose MediaTek partly because the Taiwanese firm has a strong relationship with TSMC and charges Google less per chip compared to Broadcom, the Information report added. Google spent between $6 billion (roughly Rs. 51,952 crore) and $9 billion (roughly Rs. 77,929 crore) on TPUs last year, according to research firm Omdia, based on Broadcom's target for AI semiconductor revenue last year. © Thomson Reuters 2025

Google may collaborate with MediaTek on its upcoming AI chip. What we know so far

Google is partnering with MediaTek to develop cost-effective AI chips, aiming for a release next year. Meanwhile, Google is also in discussions with Apple to introduce Rich Communication Services messaging in India, potentially revolutionizing business communications beyond traditional SMS while leveraging existing telecommunication partnerships.Google is teaming up with Taiwan's MediaTek to develop the next generation of its AI chips, Tensor Processing Units (TPUs), slated for release next year, according to a report by The Information on Monday. This partnership signals a shift in Google's chip strategy, though the company will continue its collaboration with Broadcom, which has been its exclusive AI chip development partner in recent years. Google's decision to collaborate with MediaTek may be driven by the company's lower per-chip costs compared to Broadcom, as well as its strong partnership with Taiwan Semiconductor Manufacturing Company (TSMC), according to The Information. Alongside its work with MediaTek, Google also designs its own AI server chips, which support internal research and development while being offered to cloud customers. This approach allows Google to stay competitive in the AI market by reducing its dependence on Nvidia, the industry's dominant player. This shift is particularly crucial as rivals like OpenAI and Meta face soaring demand for Nvidia chips. Google introduced its sixth-generation TPU late last year, providing an alternative to Nvidia chips for both itself and its Cloud customers. The decision to partner with MediaTek is reportedly driven by its strong collaboration with TSMC, a leading chip manufacturer, and its ability to offer more cost-effective chips compared to Broadcom. Research firm Omdia estimates that Google spent between $6 billion and $9 billion on TPUs last year, based on Broadcom's AI semiconductor revenue targets. Google is reportedly in talks with Apple to bring Rich Communication Services (RCS) messaging to Indian users this year. This could extend Apple's exclusive blue bubble messaging to Android devices, disrupting India's business communication landscape, which has relied on SMS while WhatsApp gains traction. Apple may collaborate with Google in India, following Google's existing partnerships with Vodafone Idea and Reliance Jio. The backend integration with telecom providers is unlikely to affect user experience, as RCS messages will still work over mobile data or Wi-Fi. However, this move could significantly impact A2P (application-to-person) messaging, widely used by banks and e-commerce platforms for official communications.

Google Is Teaming Up With MediaTek For Its Seventh-Generation AI Chip, Shifting From Broadcom To Cut Costs And Boost Efficiency

Google has traditionally partnered up with Broadcom to design its Tensor Processing Units (TPU) which are basically specialized AI chips utilized in data centers and for cloud computing. The AI accelerator chips are being used by its parent company, Alphabet, and are not the same as Tensor Gx processors used for Pixel phones. This is about to change as per a new report, for it is claimed that the tech giant is now switching from Broadcom to MediaTek to design its seventh-gen TPU, which could possibly be an attempt to improve efficiency and cut down on costs. As per a report by The Information via Reuters, Google is set to collaborate with Taiwan's MediaTek for the next generation of its TPUs, which will hit production next year. This could be possibly due to MediaTek's strong ties with TSMC and how, in comparison to Broadcom, it can offer cost-effective production. Broadcom has remained the tech giant's exclusive AI chip partner for design, but moving away from it could be part of Google's effort to reduce reliance on third-party chipmakers like NVIDIA for AI computing. Despite reports claiming that Google is moving away to a new design partner, MediaTek, it is important to highlight how the company is not compromising its ties with Broadcom completely and would still be depending on it for certain designs and the transitional phase. The strategic shift to MediaTek is primarily influenced by its ability to negotiate better manufacturing costs than Broadcom, given its strong ties with the world's leading chip foundry, TSMC. Google spent up to $9 billion on TPU last year, and saving up even a small fraction of the cost per chip would translate to billions in savings. This switch would also help Google with its other goal, which is to get more control over the chip design and reduce reliance on third-party solutions. By partnering up with MediaTek, the company could have more influence on the TPU architecture than otherwise, and it could play a significant role in designing its own chips. OpenAI and Meta are heavily reliant on Nvidia's GPUs for training and running their AI models, and in case of any shortages, the rival companies could be at a major disadvantage. Google's TPU strategy already puts it in a better position given how it has a self-sufficient AI hardware ecosystem, and, with the help of improved efficiency and lower cost by handing over some of the design to MediaTek, the company could have a major edge over the competition. Google, by investing heavily in its own AI chips and exploring diverse design partners, is avoiding any potential supply bottlenecks and strengthening its AI infrastructure.

Google eyeing MediaTek for AI TPU manufacturing: Report

Google is reportedly preparing to partner with Taiwan's MediaTek to develop its next-generation AI chips, known as Tensor Processing Units (TPUs), set to launch in 2026, according to The Information. Sources close to the project confirmed the collaboration. The report noted that Google is not ending its long-standing partnership with Broadcom, the San Jose-based chip designer. A Broadcom employee emphasized that the relationship remains intact. Similar to Nvidia, Google designs its own AI server chips for internal research and rents them to cloud customers. The report stated this strategy reduces Google's dependence on Nvidia's highly sought-after chips, offering a competitive edge in the AI race. Rivals like Microsoft-backed OpenAI and Meta Platforms have seen surging demand for Nvidia processors. In late 2024, Google unveiled its sixth-generation TPU, providing an alternative to Nvidia's industry-leading chips for itself and its cloud users, the report added. Google opted for MediaTek due to its strong ties with TSMC, a top chip manufacturer, and lower per-chip costs compared to Broadcom, the report explained. Research firm Omdia estimates Google spent $6 billion to $9 billion on TPUs in 2024, based on Broadcom's AI semiconductor revenue targets for that year. The report clarified that these TPUs differ from the Tensor chips in Google's Pixel smartphones and are built to power Google's global server network. By crafting its own chips, Google believes it gains a key advantage in AI innovation, especially as Nvidia's chips and GPUs remain scarce. Experts observed this move strengthens Google Cloud's standing against competitors like AWS.