Breaking news

Meta’s Bold Move: Testing In-House Chips For AI Training

Meta is reportedly piloting its own AI training chips, aiming to cut reliance on Nvidia and other hardware suppliers.

The social media giant has partnered with Taiwan-based TSMC to manufacture the custom-designed chip, optimized for AI-specific workloads. The company has initiated a small-scale deployment and, if successful, plans to ramp up production.

While Meta has previously developed AI chips, they were limited to running models rather than training them. Past chip projects have faced setbacks, with some being scrapped after failing to meet internal benchmarks.

Meta’s AI ambitions come with massive costs—this year alone, it expects to spend $65 billion in capital expenditures, much of it on Nvidia GPUs. Even a modest shift to proprietary chips could yield substantial savings and mark a major step toward greater self-sufficiency in AI infrastructure.

AI’s Economic Benefits Surpass Emissions Concerns According to IMF

The International Monetary Fund (IMF) has recently highlighted the potential economic benefits of artificial intelligence (AI), projecting a global output boost of approximately 0.5% per year from 2025 to 2030. This growth is expected to surpass the environmental costs associated with higher carbon emissions from AI-driven data centers.

The report, showcased at the IMF’s spring meeting, emphasizes the need for equitable distribution of these economic gains while managing the adverse effects on our climate. The forecast indicates that AI’s contribution to GDP growth will outweigh the financial impacts of emissions, though it points out the necessity for policymakers and businesses to mitigate societal costs.

Energy Demands and Environmental Footprint

AI is set to escalate global electricity demand, potentially reaching 1,500 terawatt-hours (TWh) by 2030, mirroring the energy consumption of countries like India today.

The increasing demand for data processing capacity could result in higher greenhouse gas emissions, but the AI industry aims to offset these with advancements in renewable energy technologies.

AI: A Driver for Energy Efficiency?

Analysts suggest that AI could potentially reduce carbon emissions through improved energy efficiency, fostering advancements in low-carbon technologies across sectors such as power, food, and transport. Grantham Research Institute stresses the significance of strategic action from governments and industries to facilitate this transition.

The role of AI in the global economy continues to evolve, stirring debates not only about its economic potential but also its environmental impact.

The Future Forbes Realty Global Properties

Become a Speaker

Become a Speaker

Become a Partner

Subscribe for our weekly newsletter