Meta, the company behind Facebook, Instagram, and WhatsApp, just announced a 19% jump in revenue for the third quarter of 2024, hitting $40.6 billion. Mark Zuckerberg pointed to the progress in artificial intelligence as a big factor.
Zuckerberg emphasized, “We had a good quarter driven by AI progress across our apps and business. We also have strong momentum with Meta AI, Llama adoption, and AI-powered glasses.”
During the earnings call, Chad Heaton, Meta’s vice president of finance, faced questions about the company’s capital expenditures. Meta’s costs and expenses rose by 15%, and they reported $9.2 billion in CapEx. Heaton clarified that the company is expanding its datacentre and server capacity to enhance both AI and other services.
He explained, “We’re still working through our 2025 infrastructure plan. At this point, we expect total infrastructure spending on non-gen AI, GenAI, and core AI to increase in 2025.”
He pointed out that they are focusing on servers, with projections showing increased spending on datacentres and network equipment as they expand their IT infrastructure. There’s also a plan to support greater user engagement beyond AI while keeping existing servers updated.
Heaton mentioned the transition into the core construction phase for smaller, higher-density datacentres. “We’re building higher-capacity networks to accommodate the growth in GenAI and core AI-related traffic in 2025. We’re also investing in fibre to manage future cross-region training traffic.”
CFO Susan Li previously discussed extending the life of datacentre servers to save costs. However, when asked if that would continue, she said, “We don’t currently plan to extend the useful lives of our servers. The ongoing performance gains from new GPU-based servers mean we want to maximize capacity with the latest technology.”
Li also highlighted the focus on optimizing recommendation systems to boost efficiency. She noted, “We completed infrastructure efficiency projects that consolidated some of our existing recommendation systems, allowing us to focus improvements on fewer models.”
Looking at ROI, she mentioned, “Near-term opportunities really center around relevance gains as we scale our models for content recommendation.” This includes increasing model complexity, which lets Meta leverage existing data with more computing power. They also plan to expand data sets for AI, enabling the algorithms to learn from even more information.