AMD has hit a pretty big milestone in the server world: its data centre revenue has reportedly overtaken Intel’s for the first time in a first quarter.
According to DigiTimes, AMD’s EPYC server CPUs helped the company move ahead of Intel in Q1 data centre revenue. AMD had already been doing better than Intel in the data centre segment since Q3 2025, but this latest result is still a major symbolic win. For decades, Intel was basically the default name in server CPUs. Now AMD is not just catching up — it is leading in one of tech’s most important battlegrounds.
Why CPUs are suddenly hot again
The interesting part is why this is happening. The AI boom started with everyone fighting for GPUs, especially for training massive AI models. But as AI moves deeper into inference and agentic AI workloads — basically AI systems doing more tasks, more often, with more compute behind them — CPUs are becoming just as important again.
The report says data centres are shifting their CPU-to-GPU planning fast. Previously, a system could be designed around something like one CPU for eight GPUs. That ratio has moved closer to one CPU for four GPUs, and some future setups may head toward one CPU per GPU.
That is a huge change. It means hyperscalers and AI companies are not only hunting GPUs anymore; they also need tons of CPUs to keep those systems fed, coordinated, and responsive.
Supply is the real headache
Demand is not the problem. Supply is.
The article points out that both AMD and Intel could likely be earning even more if the chip supply chain was not so tight. TSMC, which manufactures many advanced chips and is a key partner for AMD, is also dealing with capacity pressure. That is one reason AMD is reportedly looking at Samsung as an additional source for chip supply.
Some AI companies are also exploring custom CPU designs or asking for extra CPU capacity because existing chipmakers cannot fully keep up. This is not just an AMD-versus-Intel story anymore. It is a full supply chain fight involving foundries, packaging, server makers, cloud providers, and AI firms.
Why Malaysia and SEA should care
For Malaysian and SEA readers, this matters more than it might seem. We are using more cloud gaming, AI tools, content creation platforms, business automation, and online services every year. All of that runs on data centres somewhere.
Malaysia and Singapore are already important data centre locations in the region, and demand for AI compute could make infrastructure even more valuable here. If CPU supply stays tight, cloud costs may remain high, AI services may scale slower, and smaller regional startups could feel the squeeze before the giant US and China players do.
For gamers, this is not directly about FPS today. But indirectly? It affects game streaming, backend servers, matchmaking, anti-cheat systems, live service analytics, and the AI tools studios use to build games faster. More powerful and available data centre hardware means better services — assuming companies do not pass every extra cost back to users, lah.
The next CPU war is already loading
AMD is preparing future server platforms such as Venice and the AI-focused Verano, while Intel is working on 18A-based Diamond Rapids and the later Coral Rapids platform. Meanwhile, Arm-based competition is also heating up, with NVIDIA’s Vera tied to its Rubin platform and Arm seeing stronger expectations around AGI CPUs.
So yes, AMD beating Intel in Q1 data centre revenue is the headline. But the bigger story is that CPUs are no longer the boring sidekick in the AI era. They are becoming a core weapon again — and the companies that can ship enough of them may shape the next wave of cloud and AI infrastructure.
Source: Wccftech Gaming