REVIEWS / LAPTOPS / OWNER INSIGHTS
🦉 WE READ 620 OWNER COMMENTS
Apple M3: what owners actually say
Owners praise the M3's performance-per-watt but are relentless about 8GB RAM being unacceptable at this price
What owners complain about
- 8GB base RAM is insulting in 2023 COMMON
Numerous commenters call 8GB on machines costing $1,599+ 'insane' and 'wild.' One notes their four-year-old phone has more RAM. Apple's claim that 8GB on M3 equals 16GB on Intel is called 'gaslighting' and 'demonstrably false.'
- RAM upgrades are price-gouged COMMON
Apple charges $200 to go from 8GB to 16GB. One commenter notes they bought 32GB of DDR5 for a Framework laptop for $95. Another quips Apple gets their RAM 'from 2008' at '$50/GB.'
- Fanless machines still overheat SOME
The top-voted comment expresses surprise that a fanless computer overheats. One commenter says they expected throttling or shutdown before reaching such temperatures, implying it got dangerously hot.
- Base models depreciate badly FEW
A HN commenter notes base configuration Macs flood the used market and take a hard depreciation hit, while custom build-to-order machines hold value better because they're harder to find.
- Prioritizing looks over function persists SOME
Even after Jony Ive's departure, commenters feel Apple still prioritizes aesthetics over thermal and functional design. One sarcastically calls a machine the 'MacBook Semi-Pro' and another the 'MacBook not enough air.'
What owners love
- Performance per watt is class-leading
Multiple HN commenters argue the M-series efficiency is so good it would be viable in data centers, where electricity and cooling are the biggest ongoing costs. This is seen as Apple's genuine competitive advantage.
- Excellent processor held back by other decisions
Even harsh critics concede the SoC itself is excellent. One commenter writes, 'they have an excellent processor. Then they pair it with anemic RAM.' The silicon is not the problem.
- Neural Engine gives real edge-inference value
A commenter with years of compiler experience praises Apple's NPU/Neural Engine as 'secret sauce' that makes their hardware competitive for endpoint and edge inference compared to standard x86 CPUs.
- M3 Ultra offers unique value for certain workloads
One commenter who evaluated enterprise alternatives calls the M3 Ultra 'uniquely competent and highly affordable in a sea of enterprise technology that is decidedly not,' finding nothing else in its category matches its practical performance.
Surprising patterns
- Multiple highly-upvoted commenters argue Apple's NPU bet was strategically wrong — they assumed AI inference compute costs would shrink, but LLMs scaled UP instead, leaving Apple's GPU architecture behind for modern AI workloads.
- Commenters note that the people most affected by the 8GB RAM problem — prospective base-model buyers — are 'utterly oblivious' to the debate, comparing the discussion to 'body builders arguing about how obese people should stop eating shit.'
- A technical commenter notes Apple does not disclose GPU clock frequency for the M3 Ultra's 80 graphics cores, making direct NVIDIA comparisons difficult and suggesting Apple's FP32 execution unit count may be misleading without clock context.
WHO SHOULD SKIP IT
Anyone whose workload requires significant memory (AI inference, large models, heavy multitasking) should avoid base configurations, as 8GB is a hard bottleneck that Apple charges a steep premium to escape.
Synthesised from 620 real owner comments across 5 platforms. Every point is grounded in the comments — no marketing, no AI guessing. How we do it →