Six hundred dollars separates these two handhelds, and according to ETA PRIME's testing, the extra money bought the slower chip. That is the kind of result a spec sheet does not advertise, so somebody had to put a wattmeter on it, which is exactly what ETA PRIME did with the ROG Xbox Ally X20 and the MSI Claw 8 EX AI+.

The setup is as close to a controlled experiment as handheld reviewing gets: the Ally X20's AMD Ryzen AI Z2 Extreme against the Claw's Intel Arc G3 Extreme, both capped at identical 35W, 25W, and 17W power limits, both running Geekbench 7 and 3DMark Time Spy before a single frame of a real game loaded. Upscaling got matched too, XeSS Balanced on the Intel side against FSR Performance on the AMD side, which is the only honest way to compare two different vendors' idea of 'optimized' resolution scaling rather than just eyeballing whichever one looks sharper in a thumbnail.

The game suite is where it gets interesting: Cyberpunk 2077, Shadow of the Tomb Raider, Red Dead Redemption 2, Forza Horizon 6, and Horizon Zero Dawn Remastered, run across all three power states and also at native 1080p. ETA PRIME's numbers have the Arc G3 Extreme ahead across that stack, enough to title the video the way the channel did, though the margins and which titles favor which chip at which TDP are the kind of detail you actually have to watch unfold rather than take on faith from a thumbnail claim. A battery test against matched 80Wh capacity rounds out the comparison, because a handheld that wins benchmarks tethered to a wall outlet has won nothing a commuter cares about.

Then there is the number that actually stings: $1,299 for the ROG Xbox Ally X20 against $1,799 for the MSI Claw 8 EX AI+, per ETA PRIME's breakdown. A five-hundred-dollar premium is supposed to buy you the winning chip, not get handed to the one that lost most of the test suite. If that pricing holds up at retail, Intel's 'Extreme' branding in the Arc G3 did its job in this one shootout and AMD's silicon still ships in the cheaper box, which is a strange kind of victory for a platform that is supposed to be the budget pick.

Run your own TDP sweep before buying either one, because a 17W cap on a couch is a different machine than a 35W cap plugged into the wall, and ETA PRIME's numbers are one channel's test bench, not a universal constant. The chip war in handhelds just got a scoreboard. The price war on top of it is still unsettled, and that's the number that'll actually decide which of these sits in more backpacks.