Re: What did you expect?
Partially that it's not very useful to compare power density when thinking about heat dissipation (the sun's core has less power density than a mouse, c.f. XKCD https://what-if.xkcd.com/148/) while ballpark inaccuracies become more important when you're dealing with things involving exponentials (as heat conductance does).
Having just looked at a kettle, ~20cm cylinder, somewhere 20-30cm high, 3kW, about 1.6 W cm^-2 to air assuming the base is well insulated, and not accounting for the fairly uneven heat distribution within it if not full.
Now take into account that the 0.5W cm^-2 figure is assuming the full display is lit at a power intensity to match sunlight. But the luminous efficiency of the OLED display is likely to be higher than sunlight, which has a lot of its power in infra-red and some in UV, rather than tuned to human vision. Factor in also the display is unlikely to be running at max power on all pixels * colours (or there'd be no contrast), so for normal use there's some disregarded multipliers in there. What's important is making sure generated heat gets evened out, avoiding hotspots. So you've got a recipe for something that gets warm, but to a temperature above ambient that's a fraction that of boiling.
Biting the hand that feeds IT
