Intel spent months telling anyone who would listen that Arrow Lake would fix what Alder Lake and Raptor Lake broke.
The pitch was simple: desktop chips that finally stopped guzzling power, ran cooler, and didn't need a small power plant to hit their boost clocks.
Reviewers got their hands on the Core Ultra 200S lineup in late October, and the numbers that came back weren't the coronation Intel scripted.
Gaming benchmarks landed first, and they hurt.
Across a spread of titles, the new flagship traded blows with Intel's own previous-generation parts, sometimes losing to them outright.
For a chip family marketed as a generational leap, matching your own last-gen silicon in the one workload enthusiasts care about most is not a leap.
It's a lateral move dressed up in a new socket.
The power story is where things get genuinely strange.
Arrow Lake's headline feature was efficiency, and the chips do sip less juice under load.
But independent testers found the memory controller and interconnect running at ratios that leave performance on the table, and Intel's own launch-day guidance quietly recommended tuning that most buyers will never touch.
A platform that needs a forum thread to unlock its advertised behavior isn't a platform that shipped finished.
Intel pushed microcode and BIOS updates within weeks of launch, with motherboard vendors scrambling to ship firmware that restored lost frames.
When a CPU needs an emergency software fix to close a double-digit performance gap, that's not a driver quirk.
That's a product that left the factory before it was done.
Under sustained multi-threaded workloads, Arrow Lake is competent, cool, and cheap to run.
For someone building a workstation that occasionally games, the math can work.
But the enthusiast desktop crowd, the people who camp on launch-day reviews and argue about ring bus latency, got sold a narrative that the benchmarks don't back up.
The bigger tell is what Intel didn't emphasize.
The company spent its keynote on efficiency charts and AI branding, not on frame rates.
When a chipmaker leads with power draw instead of performance, it's usually because performance isn't the story it wants told.
Arrow Lake will likely improve with firmware and mature pricing, the way most rocky launches do.
But buying a CPU at full launch price to beta test Intel's microcode roadmap is a bad trade.
Final Thoughts
Let the early adopters absorb the pain, then buy the chip that eventually shows up.