Google Develops Frozen v2 AI Chip, Expected to Launch in 2028, Energy Efficiency Improved Up to 10 Times

Alphabet, the parent company of Google, is developing a new AI server chip codenamed "Frozen v2" to improve the operational efficiency of its self-developed Gemini model. According to The Information, citing anonymous sources, the chip is expected to be launched in 2028, and its efficiency in generating tokens per unit of energy could be 6 to 10 times that of Google's current AI chips.

gemini

In response to the report, Google told TechCrunch that the company continues to explore and test innovative technologies to enhance system performance and efficiency, but not all R&D projects will eventually reach mass production. Google emphasized that the "full-stack development" approach, which combines hardware and software design, can create more integrated computing systems optimized for real-world AI workloads.

In recent years, as generative AI applications have rapidly expanded, global AI compute demand has continued to grow, prompting tech companies to increase investment in custom chips to improve model performance and reduce reliance on external suppliers. NVIDIA has long dominated the AI chip market with its GPU technology, prompting AI companies like OpenAI and Anthropic to explore their own chip routes. In June this year, OpenAI released its first custom inference chip, "Jalapeño"; recently, Anthropic was also reported to be negotiating chip manufacturing cooperation with Samsung.

Alphabet previously announced that it plans to invest between $18 billion and $19 billion to advance its artificial intelligence strategy. As the scale of AI infrastructure investments grows, chip efficiency has become an important metric for evaluating return on investment. After the release of information about Frozen v2, market confidence increased, and Alphabet's stock rose by about 3% in the early trading session on the day of the report, with investors focusing on whether its AI investments can translate into long-term competitive advantages.

As AI model sizes continue to grow, custom chips have become an important strategic direction for tech companies to build compute ecosystems and enhance cost control capabilities. Google's latest move also reflects that competition in the AI industry is expanding from model capabilities to chip, infrastructure, and full-stack technology system competition.

Google Confirms Gemini Powers New Version of Siri: Apple's Context-Aware Function Scheduled for Release in 2026

Google and Apple have reached a cloud service partnership, developing the next-generation Apple Foundation devices based on the Gemini model, with plans to launch a new Siri in 2026. The collaboration aims to combine Gemini's multimodal processing capabilities with Apple's proprietary cloud architecture, enhancing Siri's performance in handling complex instructions and cross-app recognition.

Breaking Dependence on NVIDIA! Apple Launches Its First Self-Developed AI Server Chip Project, Baltra

Apple is accelerating the development of its self-developed AI server chip, codenamed Baltra, which is expected to be put into use in 2027, aiming to reduce reliance on NVIDIA. The project, in collaboration with Broadcom, focuses on core network transmission technology, marking Apple's deepening vertical integration strategy in the field of computing infrastructure.

AI Developers Rejoice! Google Gemini 1.5 Upgrade: Performance Soars, Prices Halved

Google today announced the launch of the newly upgraded Gemini model series, which includes Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002. This update not only significantly enhances performance but also brings surprising price discounts, undoubtedly creating a buzz in the AI development community. Firstly, the most eye-catching aspect is the substantial price reduction. The usage cost of the new models has been halved, with a drop of over 50%. At the same time, performance has seen remarkable improvements.

Google Announces 2024 I/O Developer Conference to Be Held on May 14 Focused on Gemini Model and Android 15

Google will hold the 2024 I/O Developer Conference online on May 14, with the keynote speech streamed over the internet to attract a broader developer audience. Google will showcase new features of the Gemini model and Android 15, including upgrades to Gmail and Google Photos, as well as a potential mid-range phone launch. The Gemini model will be embedded in devices to compete with OpenAI's Sora, demonstrating powerful multimodal mixture of experts model capabilities. Google releases Gemini 1.5.

Google Chrome Browser Introduces 'Help Me Write' AI Writing Tool

Google has launched a new feature 'Help Me Write', an upgraded AI writing tool. The new feature includes generative AI to assist users in composing emails with various tone options. Utilizing the Gemini model, it helps users draft texts, comments, inquiries, and more. The 'Help Me Write' feature is being promoted to Chrome users globally.