I haven't done any thermodynamic calculations on the problem but I really have used a few passively cooled computers for years now. I have roughly the same indoor temperature as you, though cooler in winter time, because heating is expensive. No AC in the summer so can indeed be pretty warm some days.


So far I haven't experienced any overheating problems but I'm not taxing the systems so much. I don't write code fast enough to make the PC burn.


In my case the systems are low end low power (and low electricity bill) processors. The last one a J 5005 Pentium system on a chip and before that J 1900 Celeron. Integrated graphics. Those systems are not producing sooo much waste heat, which simplifies the passive cooling. The Celeron system is a backbone from shuttle.eu with some heat-pipe / heatsink combination. That one still runs but is annoyingly slow nowadays. The newer Pentium was bought as a NUC and unfortunately I damaged it a bit while removing the case. Anyway.. most of it still works and I bought a case from AKASA constructed for passive cooling but lacking heat-pipes, it seemes. So now the whole case does the job as a heatsink. The system is running well. The case do tend to get warm as it's supposed to, but not dramatically. I don't think it's hot enough to fry eggs on it.


All in all I think you could realistically get a fanless computer for home use if you one day decide to get a new one. That is if you can make do with low end computer. There are no doubt fanless high end computers too but I haven't explored that market much. A Pentium is sufficient for me. Most of the time.