Zuckerberg states that Meta will have 1.3 million GPUs for AI by the end of 2025

AIbase基地

Published inAI News · 4 min read · Jan 25, 2025

136

Meta CEO Mark Zuckerberg announced in a recent Facebook post that the company plans to significantly increase its capital expenditure in 2025, aiming to maintain its leading position in the fierce competition of artificial intelligence.

Zuckerberg stated that Meta expects its capital expenditure in 2025 to reach between $60 billion and $80 billion, primarily for building data centers and expanding its AI development team. This budget range is nearly double the $35 billion to $40 billion spent by Meta last year.

Chip Technology (1)

Image Source Note: Image generated by AI, image licensed from Midjourney

Zuckerberg also revealed that Meta plans to roll out approximately 1 gigawatt of computing power this year, which is equivalent to the electricity consumption of 750,000 average households. Additionally, by the end of the year, the company's data centers are expected to have over 1.3 million GPUs (graphics processing units), greatly enhancing Meta's computing capabilities in the AI field.

As competition in the AI sector intensifies, Meta's investment plan becomes particularly crucial. Currently, many competitors are also increasing their investments in infrastructure. For example, Microsoft plans to invest $80 billion in AI data centers by 2025, while OpenAI is participating in a joint investment project called Stargate, which could provide it with data center resources worth hundreds of billions of dollars.

Meta's significant investment in AI not only reflects the company's commitment to technological development but also indicates its competitive stance in the industry. Zuckerberg emphasized in his post that in the coming years, Meta will focus on building a more robust AI infrastructure to support its competitiveness in the global market.

Key Points:
🔹 Meta plans to increase capital expenditure to $60 billion to $80 billion in 2025, primarily for AI and data centers.
🔹 Approximately 1 gigawatt of computing power is expected to be online this year, equivalent to the electricity consumption of 750,000 households.
🔹 Competitors are also ramping up investments, with Microsoft planning to invest $80 billion, and projects involving OpenAI potentially bringing in hundreds of billions of resources.

UK Proposes Law to Ease AI Training: Strong Opposition from Music Industry Calling for 'Real Control'

The UK government is considering amending laws to allow AI companies to use creators' online content to train their models unless creators explicitly opt out of the system. This move has sparked strong opposition from the music industry, with former Beatles member Paul McCartney openly expressing his concerns, warning that it could lead to a 'Wild West' scenario where creative works lose the copyright protections they deserve. In an interview with the BBC, McCartney urged the government to reconsider this approach and emphasized the potential threat it poses to emerging artists: 'Young boys, girls...'

DeepSeek AI Assistant Rises to the Top of the Apple Store: Low Cost, High Efficiency, China's AI Ascendancy Draws Global Attention

Chinese artificial intelligence company DeepSeek recently released the open version of its inference model R1, which has quickly sparked discussions in the tech community. Its remarkable breakthrough achievements left venture capitalist Marc Andreessen in awe, describing it as 'one of the most astounding and impressive breakthroughs I've ever seen,' and it has demonstrated capabilities in AI benchmarking that rival or even surpass those of OpenAI's o1 model. Notably, DeepSeek claims that its model training cost is only $5.6 million, whereas leading U.S. companies require hundreds of millions.

AI Has Fatal Flaws in This Area! Experts: AI Should Not Be Used in Critical Applications

In the current technological environment, artificial intelligence (AI) has sparked widespread discussion. Eerke Boiten, a professor of cybersecurity at De Montfort University, states that existing AI systems have fundamental flaws in management and reliability, and therefore should not be used in critical applications. Professor Boiten points out that most current AI systems rely heavily on large neural networks, particularly generative AI and large language models (like ChatGPT). The workings of these systems are relatively complex, even though each neural...

Character AI Files Motion to Dismiss Lawsuit, Claims Protection Under First Amendment

Recently, the AI chatbot platform Character AI found itself in legal trouble due to a case involving teen suicide. The platform filed a motion to dismiss in the U.S. District Court for the Middle District of Florida, arguing that under the First Amendment, they should not bear responsibility for the related lawsuit. Source of the image: Image generated by AI, image licensed from service provider Midjourney. The case originated from a lawsuit filed against Character AI by Megan Garcia in October.