TrimUI Smart Pro
Not viable on this hardware, regardless of settings or custom firmware.
Tested by: Retro Handhelds Online
PlayStation 2 (Sony, 2000) tested on 11 devices, cheapest first.
8 of 11 tested handhelds run PS2 well enough to play, starting at $199.
4 of them hold full speed. Ratings come from 16 publishers who had the hardware; the comparison across all of them is ours.
Around 90% of the library runs with upscaling to 1080p and widescreen hacks enabled.
Not viable on this hardware, regardless of settings or custom firmware.
Tested by: Retro Handhelds Online
Some titles run untouched at native resolution; most popular ones mean a trip to the settings menu.
Tested by: Joey's Retro Handhelds, RetroGameTalk
More variable than GameCube. A per-title question rather than a yes.
Tested by: Retro Handhelds Online, Retro Game Corps
Around 90% of the library runs with upscaling to 1080p and widescreen hacks enabled.
Tested by: Handheld Index, Retro Handhelds Online
Runs reliably as part of the device core library.
Tested by: RetroSpecGame
Runs, but expect 30-45 fps on many titles rather than the original 60.
Tested by: DroiX, Held Games
Roughly 85-90% of the library at full speed without significant tweaking.
Tested by: Held Games, Android Authority
Full speed, including titles that were demanding on earlier Android handhelds.
Tested by: RetroResolve, AndroidAyuda
Handles the library upscaled to 1080p, without much fiddling or compromise.
Tested by: overkill.wtf, Retro Handhelds
Demanding titles run at 1.5x to 2x native resolution.
Tested by: Stuff, Retro Dodo
Runs as it would on original hardware, upscaled to 720p.
Tested by: Retro Dodo, NerdBurglars
Devices are ordered by price, not by score, because the useful question is usually the cheapest thing that clears the bar rather than the best thing available. “Playable” is where most people should stop reading upward: past that you are paying for headroom and upscaling, not for the game running.
Devices absent from this list have no published PS2 testing found. That is a gap in the data, not a verdict on the hardware.
Every rating on this page traces to one of these. The compilation is ours; the measurements are theirs.