Steam’s Hardware Survey is one of the clearest monthly snapshots of what active PC gamers are actually running. Right now, that snapshot is flashing a simple message: the old Intel-heavy default is no longer safe to assume.
In the most recent survey release, Intel remains ahead, but the lead has narrowed enough that month-to-month swings now feel meaningful. If you build, tune, or ship PC games for a living, that matters more than the fandom angle. Your performance problems tend to appear first on whatever hardware is most common.
The more important story is the compounding effect of two forces moving in opposite directions: Intel taking repeated hits to its gaming reputation, and AMD building a multi-generation identity around "the gaming CPU," especially in the midrange where most Steam users live.
This is also why you should treat sweeping claims like "gamers are voting with their wallets" as shorthand. Steam’s survey is opt-in and anonymous, so it can drift with geography, new installs, and sampling quirks, but it still tends to track real momentum over time.
If you want the exact vendor split for the current month, start with Valve’s own processor vendor table. Valve also notes on the survey overview that participation is optional, which is why the survey is best used for trends and assumptions, not as a precise market share audit.
So what changed? Part of it is simple product gravity. AMD’s X3D chips turned cache into a marketing feature that also shows up in the places gamers feel it most, smoother frame pacing, better 1% lows, and fewer ugly CPU spikes in crowded scenes. That's a hardware bias toward the workloads many modern engines create.
Intel’s side of the ledger is messier. The brand took reputational damage from high-profile instability reports across 13th and 14th gen desktop parts, and Intel has acknowledged root causes tied to elevated voltage behavior that can contribute to Vmin shift instability. That official write-up is at Intel.
Even if many users never experienced a crash, "maybe unstable" is poison in the enthusiast funnel, because enthusiasts are the people friends ask before spending $300 to $600. And once that social graph tilts, retail charts tilt with it.
Meanwhile, Valve keeps nudging PC gaming into broader living-room and handheld shapes, which changes the center of gravity for performance expectations. If you are tracking where Steam’s platform bets are going, these updates are a useful backdrop: Steam Deck 2 performance and timing and Valve’s Steam Machine-style hardware return. Linux share growth is part of the same story as well, since it affects driver paths and test matrices.
What this means for game teams is practical. Don't "optimize for AMD instruction sets." Instead, update your default assumptions about cache, core topology, and scheduling behavior, then validate with profiling.
Three concrete takeaways tend to pay off across both vendors. First, treat cache and memory locality as first-class performance constraints, because cache-heavy CPUs will mask sloppy memory behavior and non-cache-heavy CPUs will punish it. Second, invest in frame-time stability work, not just average FPS, because today’s buying decisions are heavily influenced by 1% lows and stutter. Third, test on both hybrid-core Intel systems and chiplet AMD systems, because thread scheduling, background tasks, and compile or shader workloads can interact differently with each topology.
Intel can still reverse this. Reputations swing when products swing. But in the near term, the calmer interpretation is also the most useful: the "typical Steam PC" is becoming less Intel-default each quarter, and engines that assume one vendor’s quirks are going to keep finding surprises.
