Apple Mac mini 2024 (M4) review

01 December 2024
It has always been the most accessible computer the company has offered and the bare bones nature was designed primarily to make people using Windows PCs switch over.

Sort by:

  • ?
  • Anonymous
  • vxx
  • 1 hour ago

AnonF-1006353, 3 hours agoDepending on benchmark, this is true. Passmark: 8745H: 31... morePassMark? The benchmark app that was under fire for implementing an obvious anti-AMD update to skew the results towards Intel a few years ago? Surely they are credible aren't they?

Cinebench 2024 is much more demanding than R23 (6x higher computational effort for the scene) and utilises newer instruction sets. Still, the R23 could be used as a rough reference of how each CPU performs when you don't have results from the CB 2024 on hand, although the latter is clearly a better indicator for sure.

    • F
    • AnonF-1006353
    • JBi
    • 3 hours ago

    TheSnowWinterR7, 4 hours agoClock that tea ⏰☕There's no tea to be clocked, specially if you don't have any technical details to add to the discussion

      • D
      • DoodleDan777
      • 0{j
      • 3 hours ago

      justasmile, 4 hours agoSlightly better chip performance? You're definitely jokingSorry I was wrong. Switched the scores up by accident.

      However the m4 (10c) still only outperforms everything in single core, the new 7 8745h (8845h - AI) is getting better multi scores almost everywhere, which is what I need and what I focus on but I didn't reference it last post. It does require more tdp but they're both mobile chips so it's a fair comparison.

        • F
        • AnonF-1006353
        • JBi
        • 3 hours ago

        Sam.Smythe, 3 hours agoThen ban this misinformation spreader. Keep your website co... moreIt is exacty not misinformation, what I say matches the numbers.

          • F
          • AnonF-1006353
          • JBi
          • 3 hours ago

          justasmile, 4 hours agoSlightly better chip performance? You're definitely jokingDepending on benchmark, this is true.
          Passmark:
          8745H: 31k
          M4-10: 24k

          Cinebench R23:
          15930 for the 8845HS and the 8745H should be 6% faster so 16k+
          M4-10: 13.9k

            prasad-gsma, 9 hours agoNo, the other guy got it right. I can see your entire comme... moreThen ban this misinformation spreader. Keep your website comment section clean, tidy, and free from misinformation and hate speech.

              prasad-gsma, 9 hours agoNo, the other guy got it right. I can see your entire comme... moreClock that tea ⏰☕

                DoodleDan777, 8 hours agoIt kinda sucks. It's Apple. I would atleast expect a g... moreSlightly better chip performance? You're definitely joking

                  • ?
                  • Anonymous
                  • qb{
                  • 6 hours ago

                  [deleted post]is saying, "nobody really needs to edit documents in 8k 480fps" gonna make people outraged

                    • F
                    • AnonF-1006353
                    • JBi
                    • 6 hours ago

                    prasad-gsma, 6 hours agoI can't comment on the older Cinebench R23 results as ... moreFair enough.

                    Even if I'd like to know what exactly leads to the R23 vs 2024 cinebench differences. E.g. what component is weighed more than before and if this fits the most use cases

                      AnonF-1006353, 7 hours agoI'm not fooling anyone really, while: "the M4&... moreI can't comment on the older Cinebench R23 results as none of the credible reviewers I know have tested it on the new Intel and AMD chips nor would I put much stock in a 100 point difference in that app. But if the Intel and AMD chips are actually faster then fair enough. But I do know that the M4 is almost 26% faster than the 285K or the 9950X in the latest Cinebench 2024, which is far outside of a fluke or test variation range.

                      The 8700G was used for comparison because it's a desktop part as is the M4 specifically inside the Mac mini, which has none of the thermal or power constraints that it would inside a MacBook. Had I compared against a mobile chip I'm sure someone would have asked why I was comparing desktop parts against mobile chips. This is why there were no comparisons against the Snapdragon X Elite either, even though the M4 beats it in everything except Cinebench 2024 multi-core while still consuming half as much power.

                        • F
                        • AnonF-1006353
                        • JBi
                        • 6 hours ago

                        prasad-gsma, 7 hours agoThe review features nine benchmarks, all of which are indus... moreWith biased I was mainly refering to geekbench.
                        The web results are nice but are essentially a browser engine (dependent) benchmark. While they might resemble a real world scenario, it's only valid for the specific browser and version used. We could now argue about wether we're talking about the end user experience performance or raw cpu/gpu performance, but for the latter I'd trust a simple prime number calculation benchmark like PassMark more, compared to rendering or web benchmarks where multiple variables and different weighing of certain cpu/gpu features and optimizations can totally change the result

                          • F
                          • AnonF-1006353
                          • JBi
                          • 7 hours ago

                          prasad-gsma, 9 hours agoNo, the other guy got it right. I can see your entire comme... moreI'm not fooling anyone really, while:

                          "the M4's single-core performance in Geekbench and Cinebench is chart-topping, beating out every single consumer CPU on the market, including flagship parts such as the AMD Ryzen 9 9950X and the Intel Core Ultra 9 285K."

                          Might fool some, because for Cinebench R23 results, this statement is clearly not valid, as even according to notebookcheck the 285K is in the 2350 range, compared to 2220 for the M4 (base model 10 core variant).

                          Also mentioning the 8700G is a bit misleading, when the newer 370 HX with 890M GPU exists, which is about 30% more performant compared to 780M. Same goes for the 7600X while 9600X is available.

                          I agree that Intel, as well as AMD and Qualcomm need to up their game regarding single-core performance vs power rating (or even multi). But how should they, if apple reserves all 3nm capacities to themselves, which is quite an anti-competition move in my book.

                            AnonF-1006353, 8 hours agoI said what was needed to be said. I didn't say it... moreThe review features nine benchmarks, all of which are industry standard across platforms and used by everyone, including Apple's competitors, in their first-party testing. Everything that can be run and is relevant has been included, although the list is still short considering macOS still doesn't have as many benchmarks for it as Windows. To suggest that somehow all nine results are not relevant and biased but the one cherry-picked benchmark that no one even uses anymore isn't is laughable.

                              Those prices for storage and ram upgrades i mean, just doubling it, they rip you hundreds of euros... i can see how they got to trillion dollar company.

                                • D
                                • DoodleDan777
                                • 0{j
                                • 8 hours ago

                                It kinda sucks. It's Apple. I would atleast expect a graphene cooling system from them at this point. The M4 is good but that price just doesn't match up with the competition, the um870 slim has slightly better chip performance, 32gb of ram and a Terabyte of storage at just $529.
                                So the Mac Mini is honestly useless if you don't have a portable ssd with you at all times.

                                  • F
                                  • AnonF-1006353
                                  • JBi
                                  • 8 hours ago

                                  I said what was needed to be said. I didn't say it's slow, I just said the review is not giving the full picture. But those who don't get it, can continue to hallow only biased benchmarks which originated out of an apple fan community and its corresponding limited, non-user upgradeable $600 one-trick pony.

                                    Anonymous, 9 hours agoThe 16GB/256GB base model is amazing in terms of price to p... moreYep, the base M4 model is unironically a fantastic deal performance-to-price ratio wise, and the base M4 Pro variant isn't too bad either.. but unfortunately Apple's upgrade options are beyond ridiculous at this point.

                                    For reference, upgrading the Mac Mini's SSD to 2TB would add a whopping 830 USD (equivalent) to the cost where I live. A similarly sized 2TB internal SSD with a faster read/write speed (Crucial P310 2TB NVMe M.2 Gen4 Internal 2230 SSD) is priced around 200 USD in comparison. It's pretty obvious that Apple is using a proprietary SSD standard to allow them to jack up the price however they want. Also, I know the packaged RAM is part of the design, but many Snapdragon X Elite powered laptops don't charge buyers as much with a RAM upgrade.

                                    Luckily for Mac Mini buyers, since it's desktop, if they want a higher capacity storage, they can plug in a much cheaper portable SSD to one of the TB4 ports and call it a day - It wouldn't justify Apple's pricing, but for most users they are not forced to pay the full price for the upgrade if they want more. M4 MacBook Pro, however..

                                      AnonF-1006353, 10 hours agoCinebench R23 single core: Intel Core Ultra 5 245KF: 238... moreIs there any logical reasoning behind comparing M4 to two CPUs that have PL1 of 125W?

                                        • F
                                        • AnonF-1006353
                                        • JBi
                                        • 9 hours ago

                                        prasad-gsma, 9 hours agoYeah, it's chart topping in benchmarks that are actual... moreComputers don't only exist for 3D or video rendering, I would have 0 use cases for this. I'd be much more interested in 7z archiving or some independent compiling times for example. Where passmark will be much closer to reality as it's less dependent on various optimizations. That much for actual relevancy.

                                          1 of 2