Google's TPU and Gemini leap

Google's release of its Gemini 3 LLM in November, trained primarily on the company's in-house TPU chips, is performing at or above the level of OpenAI's ChatGPT.

Google's release of its Gemini 3 LLM in November—trained primarily on the company's in-house TPU chips— is performing at or above the level of OpenAI's ChatGPT.

This development has become a catalyst for Google's broader strategic push amidst an intensifying AI arms race. The company is building a vertically integrated stack, one that it hopes will form a defensible moat in computing power.

DIGITMES Asia provides basic users with unlimited access to toll-free articles. Sign up for free subscription now to continue your reading!

Subscribe Now

IN THE NEWS

Monday 1 December 2025

Research Insight: Gemini threatens OpenAI's AI crown: Ecosystem advantage signals strategic shift in tech landscape

Google has launched Gemini 3, a new generative AI model that has quickly outperformed competing models, including OpenAI's recently released GPT-5.1, across most AI benchmark leaderboards...

Saturday 29 November 2025

GPU and TPU dispute illustrates zero-sum fallacy

The AI chip market is witnessing intensifying competition as Google's tensor processing units (TPUs) gain momentum alongside Nvidia's dominant graphics processing units (GPUs). With...

Friday 28 November 2025

Yesterday AOL, today OpenAI: Google's vertical integration threatens to repeat history

In the competitive realm of artificial intelligence computing power, Nvidia currently leads the market, but Google is emerging as a formidable contender with its self-developed TPUs...

Thursday 27 November 2025

Google's Gemini 3 launch marks founder Brin's return to active AI strategy

Google has made a decisive comeback in the artificial intelligence sector with the recent launch of its upgraded Gemini 3 large language model and self-developed Tensor Processing...

Thursday 27 November 2025

Inventec gains from Google TPU drive as cloud firms seek Nvidia alternatives

Google's renewed focus on custom Tensor Processing Units (TPUs) for artificial intelligence is driving up demand for its server hardware, giving Taiwan's Inventec new momentum and...

Thursday 27 November 2025

Samsung poised for memory and foundry gains as Google TPU push accelerates

Google's push to expand its in-house Tensor Processing Unit (TPU) platform is drawing fresh attention across the semiconductor sector, and analysts say Samsung Electronics could emerge...

Thursday 27 November 2025

Google pushes TPU-OCS architecture as Gemini 3 aims at ChatGPT

Google's release of its Gemini 3 large language model (LLM) in November—trained primarily on the company's in-house TPU chips and...

Wednesday 26 November 2025

Nvidia defends GPU lead as Google TPU gains traction in AI chip market

Nvidia has asserted that its graphics processing unit (GPU) platform remains a full generation ahead of its competitors, responding to increased attention on Google's Tensor Processing...

Wednesday 26 November 2025

Google steps up in ASIC war, but brings mixed implications for MediaTek and Broadcom

Google has been on a streak of breakthroughs in AI. Gemini 3 has been outperforming major AI models, while it has also received strong engineering feedback on Antigravity. Its already...

Tuesday 25 November 2025

Meta reportedly eyes Google TPU adoption, intensifying pressure on Nvidia

Meta is reportedly negotiating to adopt Google's tensor processing units (TPUs) in its data centers starting in 2027, with plans to lease TPUs from Google Cloud as early as 2026....

Tuesday 25 November 2025

Google's Nano Banana Pro boosts text-accurate AI imagery but still shows reasoning limits

Google has introduced Nano Banana Pro, a Gemini 3 Pro–based image model that improves multilingual text rendering and visual consistency while enabling more reliable infographics...