Close Menu

    Stay Ahead with Exclusive Updates!

    Enter your email below and be the first to know what’s happening in the ever-evolving world of technology!

    What's Hot

    Grid-Responsive AI: How Nvidia Plans to Turn Data Centers Into Power Assets with Emerald AI

    April 16, 2026

    Cyber Retaliation: How Iran-Linked Hackers Paralyzed Medical Giant Stryker

    April 16, 2026

    The Trillion-Dollar Exit: Why a SpaceX IPO Would Reshape the Space Economy

    April 15, 2026
    Facebook X (Twitter) Instagram
    Facebook X (Twitter)
    PhronewsPhronews
    • Home
    • Big Tech & Startups

      The Trillion-Dollar Exit: Why a SpaceX IPO Would Reshape the Space Economy

      April 15, 2026

      The Sacramento Blueprint: How California is Writing the U.S. AI Rulebook

      April 14, 2026

      Silicon Sovereignty: Microsoft’s $10B Play to Secure Japan’s AI Future

      April 14, 2026

      The Rule-Breaker: Why DeepSeek V4 is China’s Defiant Bid for AI Supremacy

      April 13, 2026

      Scheduling Monopoly? Why Nvidia’s SchedMD Deal Alarms Open-Source Devs

      April 13, 2026
    • Crypto

      Quantum Computing Advances Force Coinbase and Institutional Custodians to Rethink Crypto Security

      March 8, 2026

      AI Assisted Hacking Groups Target Crypto Firms With Multi-Layered Social Engineering

      February 18, 2026

      Global Crypto Regulations Expand as 2026 Begins With New Data Collection Frameworks and National Laws

      January 16, 2026

      Coinbase Bets on Stablecoin and On-Chain Growth as Key Market Drivers in 2026 Strategy

      January 10, 2026

      Tether Faces Ongoing Transparency Questions and Reserve Scrutiny Amid Massive Bitcoin Accumulation

      January 5, 2026
    • Gadgets & Smart Tech
      Featured

      AirPods Max 2: USB-C, Live Translation, and the H2 Upgrade

      By preciousMarch 26, 2026
      Recent

      AirPods Max 2: USB-C, Live Translation, and the H2 Upgrade

      March 26, 2026

      How ABB and Nvidia are Perfecting Industrial Robotics using AI Simulation

      March 20, 2026

      Neura Robotics Reaches €4B Valuation With Tether Backing

      March 12, 2026
    • Cybersecurity & Online Safety

      Cyber Retaliation: How Iran-Linked Hackers Paralyzed Medical Giant Stryker

      April 16, 2026

      Your Company Could Be Iran’s Next Target: What U.S. Tech Firms Need to Do Right Now

      April 6, 2026

      Google Is Warning Us About The Encryption Protecting Your Data Today. It May Not Survive Quantum Computing

      April 5, 2026

      Accenture and Anthropic Team Up on AI-powered Cybersecurity

      April 4, 2026

      Your BVN, Passport, and Bank Account May Already Be on the Dark Web. What Every Nigerian Must Do Right Now After the Banking Breaches

      April 4, 2026
    PhronewsPhronews
    Home»Artificial Intelligence & The Future»Google teams up with Taiwan’s MediaTek for next AI chip
    Artificial Intelligence & The Future

    Google teams up with Taiwan’s MediaTek for next AI chip

    oluchiBy oluchiMarch 21, 2025Updated:June 12, 2025No Comments
    Facebook Twitter Pinterest LinkedIn WhatsApp Reddit Tumblr Email
    Share
    Facebook Twitter LinkedIn Pinterest Email

    Google announced a strategic partnership with Taiwanese semiconductor giant MediaTek to develop its next-generation custom AI chips, Tensor Processing Units (TPU).  On March 17, the announcement of this partnership sparked speculation about its potential to transform the semiconductor landscape.

    This partnership, however, does not imply that Google will be severing ties with Broadcom, its longtime partner in manufacturing TPUs, but instead it is a step to mitigate supply risks by having multiple suppliers and enhance the supply chain resilience.

    For the past decade, Google has been developing custom AI accelerators, the TPU, to address the growing demands of AI workloads and pave the way for multimodal AI.  Unlike general-purpose chips like the CPU, TPUs are application-specific integrated circuits (ASICs).

    The TPU’s specific design accelerates a high volume of mathematical and logical operations, thereby speeding up and improving the efficiency of machine learning.  In early 2024, Google announced ‘Trillium,’ a sixth-generation TPU, a high-performance chip specifically made for generative AI in mind.

    There is a significant distinction between Trillium and the previous generations of TPUs.  Trillium boasts a 4x increased training speed, a 3x increase in inference throughput, a 67% energy efficiency boost, a 4.7x increase in peak compute per chip, double HBM (High Bandwidth Memory) capacity and ICI (Interchip Interconnect) bandwidth, a 100k-chip network, and improved price-performance.

    These features facilitate enhanced performance across diverse AI workloads, including scaled training, LLM training (dense and Mixture of Experts (MoE)), enhancing inference and scheduling, handling embedding-intensive models, and delivering price-performance for both training and inference.

    Trillium was used to train Google’s Gemini model 2.0 and serves as a key component of Google’s AI Hypercomputer, a system that features over 100,000 Trillium chips via a Jupiter network fabric delivering 13 petabits/sec of bandwidth capable of scaling a single distributed training job to hundreds of thousands of accelerators.

    Although Google does not sell the Trillium chips directly, they make it available via Google Cloud, making the infrastructure more accessible for a wide range of applications. Trillium  TPU was made generally available to the public in December 2024.

    The TPU, Trillium, provides an alternative to Nvidia chips for Google and its Cloud customers.  According to Economic Times, it is speculated that the decision of Google to partner with MediaTek is driven by its strong collaboration with TSMC (Taiwan Semiconductor Manufacturing Company) and MediaTek’s ability to offer more cost-effective chips compared to Broadcom.

    This partnership also has the potential to affect the current landscape of semiconductor market competition. Nvidia chips currently hold between 70% and 95% of market shares for artificial intelligence chips.  Google’s partnership with MediaTek for its next generation of TPU chips would earn it a significant percentage of market shares, as not only will it be cost efficient, but it will also have the scaling power and efficiency to be a suitable alternative to Nvidia chips.

    The partnership will also help Google enhance AI capabilities, reduce costs, reduce reliance on a single supplier, gain greater control over its chip supply, and result in Google strengthening its position in the competitive AI industry.

    AI accelerators AI chips AI development AI Hypercomputer AI industry competition AI infrastructure AI innovation AI workloads application-specific integrated circuits Broadcom chip alternatives chip manufacturing chip scaling cloud computing cost-effective chips custom AI chips distributed training energy efficiency generative AI Google Google AI capabilities Google Cloud Google Gemini high-performance chips LLM training machine learning performance MediaTek MediaTek chips Mixture of Experts next-generation chips Nvidia chips price-performance semiconductor industry semiconductor market competition semiconductor market share semiconductor supply risks supply chain resilience Tensor Processing Units TPU partnership Trillium TPU TSMC collaboration
    Share. Facebook Twitter Pinterest LinkedIn Tumblr Telegram Email
    oluchi
    • X (Twitter)
    • LinkedIn

    I am a content writer with over three years of experience. I specialize in creating clear, engaging, and value-driven content across diverse niches, and I’m now focused on the tech and business space. My strong research skills, paired with a natural storytelling ability, enable me to break down complex topics into compelling, reader-friendly articles. As an avid reader and music lover, I bring creativity, insight, and a sharp eye for detail to every piece I write.

    Related Posts

    Grid-Responsive AI: How Nvidia Plans to Turn Data Centers Into Power Assets with Emerald AI

    April 16, 2026

    The Trillion-Dollar Exit: Why a SpaceX IPO Would Reshape the Space Economy

    April 15, 2026

    The Sacramento Blueprint: How California is Writing the U.S. AI Rulebook

    April 14, 2026

    Comments are closed.

    Top Posts

    Coinbase responds to hack: customer impact and official statement

    May 22, 2025

    MIT Study Reveals ChatGPT Impairs Brain Activity & Thinking

    June 29, 2025

    From Ally to Adversary: What Elon Musk’s Feud with Trump Means for the EV Industry

    June 6, 2025

    Anthropic Will Use Claude User Chats For Data Training

    October 16, 2025
    Don't Miss
    Uncategorized

    Grid-Responsive AI: How Nvidia Plans to Turn Data Centers Into Power Assets with Emerald AI

    By preciousApril 16, 2026

    Data centers are power-hungry by design, but Nvidia wants them to work in the opposite…

    Cyber Retaliation: How Iran-Linked Hackers Paralyzed Medical Giant Stryker

    April 16, 2026

    The Trillion-Dollar Exit: Why a SpaceX IPO Would Reshape the Space Economy

    April 15, 2026

    The Sacramento Blueprint: How California is Writing the U.S. AI Rulebook

    April 14, 2026
    Stay In Touch
    • Facebook
    • Twitter
    About Us
    About Us

    Evolving from Phronesis News, Phronews brings deep insight and smart analysis to the world of technology. Stay informed, stay ahead, and navigate tech with wisdom.
    We're accepting new partnerships right now.

    Email Us: info@phronews.com

    Facebook X (Twitter) Pinterest YouTube
    Our Picks
    Most Popular

    Coinbase responds to hack: customer impact and official statement

    May 22, 2025

    MIT Study Reveals ChatGPT Impairs Brain Activity & Thinking

    June 29, 2025

    From Ally to Adversary: What Elon Musk’s Feud with Trump Means for the EV Industry

    June 6, 2025
    © 2025. Phronews.
    • Home
    • About Us
    • Get In Touch
    • Privacy Policy
    • Terms and Conditions

    Type above and press Enter to search. Press Esc to cancel.