Apple Silicon Mac mini Predictions

[Yebubbleman]

The one Mac model that there has been zero rumors or chatter about specifics regarding an Apple Silicon version is the Mac mini. I've seen people talk about how they want one or are looking forward to one, but I've heard zero in terms of rumors or speculation on what one might entail and when it might launch.

We know that Apple gave the Mac mini only adjustments to the base models this year (not even a true refresh) similar to the treatment of the 21.5" iMac when the 2020 27" iMac launched earlier this summer. It would also seem as though Apple will be easily able to release Apple Silicon versions of 8th Generation Intel based Macs (such as the Mac mini, the 21.5" iMac, and the current Two-Port 13" MacBook Pro) that easily outperform said 8th Generation Intel based versions with the technology present in the A12Z Bionic, let alone whatever they have in the pipeline. So, they'll be able to do it starting day 1. The question becomes "when will they do it and how will they do it?".

Personally, I'm hoping that, whatever they do, they don't reduce the number of ports from the 2018 model, past that, I'll be fine with whatever they do (especially since I'm not a regular Mac mini customer).

As for my predictions of what they'll do, it's safe to say that the storage will be fixed as it is now (and bound to the SoC the way the 2018 mini's storage is bound to the T2 chip today). I'm unsure of removable RAM. They did make a big deal about it when they launched the 2018 mini. Though it's not as user-friendly of a replacement/removal as it was on the 2010-2012 Mac minis. I'd guess that the SoCs in the mini will probably be similar in performance to whatever they put in the rumored Apple Silicon 24" iMac (if not also the Apple Silicon replacements to the MacBook Air and 4-port 13" MacBook Pros). Past that, I have no clue what they'll do for form factor. They could definitely shrink the design while still keeping the port count what it is and not hitting thermal limits. But I have no clue how or if they will do that for the first generations of Mac mini.

The fact that we haven't heard anything about the mini gives me pause. I'm not entirely sure when the next 16" MacBook Pro is coming out (and it's the only other Mac that details on the immediately next refresh aren't as clear on), but at least we have an idea that an Apple Silicon model will have been released no later than 2H 2021. There seem to be no such details on the Mac mini. The only thing that we can say for sure is that, by keeping 8th Gen Intel on the currently-sold Mac mini, Apple doesn't need to do much (or wait much longer) to be able to deliver an Apple Silicon model that handily bests it.

What say you all?

 

[casperes1996]

Considering the Mac mini houses the DTK with pretty much the same ports, thunderbolt aside, I would assume we will see a silent refresh to the Mac mini - what we'd normally call a spec bump - that just replaces the CPU from from the 8th Gen Intel chips to the T3 Mac chips (yes... I'm calling it T3. I don't particularly like the name I came up with but I couldn't think of any other I like better). Basically an A14Z with an A14X being in the iPad Pro models. No big fanfare or Haha. Just a press briefing one day out of the blue and the mini is Apple Silicon and way more powerful especially for graphics

 

[ivnj]

Just found this.

Apple Silicon Mac mini could just be mightier... and mini-er

Just how much mightier, how much mini-er, could Apple Silicon really make the Mac mini?

www.imore.com

 

[Woochoo]

Just found this.

Apple Silicon Mac mini could just be mightier... and mini-er

Just how much mightier, how much mini-er, could Apple Silicon really make the Mac mini?

www.imore.com

Actually there are full blown Ryzen computers + Vega in a SBC (Raspberry, basically) form factor, with the size of a credit card. Apple TV isn't that far from that in size, so it's plausible although having no active cooling and using the same chip as iPad Air would put it in a position where the only difference with those would be the OS, as for the similar price you'd have way more stuff packed in the iPad.

 

[sanfrancisofont1984]

Both AMD and Intel has several tiers of processors. ULV, notebook, desktop, workstation/server.

2018 mini is kinda odd that it is using desktop (rather than notebook) tier processor. And it is expensive.

I would be happy if the coming new mini offers the same horsepower as a MBP but at a lower price point.

 

[George Dawes]

I presume it will be even more mini due to the lack of heat from the new chips

Maybe it'll be appletv size ?

 

[BeatCrazy]

Wouldn't it be crazy if the iMac Pro was in the first batch of Apple Silicon Macs? Obviously the current gen is long-in-the-tooth. Not sure if an overclocked A14Z would bench better than the i9 you can get in the 2020 iMacs.

 

[theluggage]

2018 mini is kinda odd that it is using desktop (rather than notebook) tier processor. And it is expensive.

...which is a bit of a cheek considering that the desktop chips they are using are considerably cheaper than the successors to the mobile processors used in previous Minis. OK, you get good CPU performance, but the downside is that Intel don't have a "desktop" equivalent of their premium Iris/Pro/Plus/whatever integrated graphics, so the Mini is stuck with lowest-common-denominator graphics (...please don't suggest that I get a neat, tiny computer like a Mini and then tether it to a honking great eGPU chassis that adds at least $200 to the cost of a limited choice of supported GPUs, needs its own wall-wart etc.) - I'd have thought that - circa 2018 - something like the Skull Canyon NUC chipset or those Intel CPUs with in-package discrete AMD GPUs would have produced a better balance between CPU and GPU.

It sounds like Apple Silicon is going to have better integrated graphics than any Intel processor - potentially solving that problem - but it remains to be seen whether the first ASi-for-Mac chip is going to be able to compete with desktop Intel processors on the CPU front (we know that even the A12Z can give a mobile i7 a run for its money). Apple will have to cross that bridge at some stage if they're going to replace the 5k iMac and Mac Pro but I think it's a fairly safe bet that the first ASi chips will be targeted at the lower-end laptops, which should be the easiest win for ASi.

The current Mini design is definitely ripe for an update, seeing as it is still based around a case originally designed to house an optical drive and/or various combinations of spinning rust. So it has an over-abundance of internal space: the 2018 Mini used this to fit in a desktop-grade heatsink and fan. That's probably why it was chosen for the DTK too - room for a slightly kludgey small-run (for Apple) logic board, and the closest thing to a simple headless desktop Apple do (...the DTKs are going to be rounded up and shredded after a year, no point in wasting a whole laptop). So I wouldn't take the DTK as a sign of some new enthusiasm for the Mini on the part of Apple.

Personally, an ASi Mini would be my preferred choice for a first ASi Mac - partly because it would have to work alongside my existing iMac for the near future until everything I need got ported and, longer term, I really don't want another all-in-one. However the reality is that the Mini has never been a big priority to Apple and I'm really not holding my breath.

If they do upgrade the Mini - fixed SSD sucks but is probably inevitable (really don't see Apple back tracking on that) and, in any case, it's going to be proprietary ASi-driven SSD. They could add a standard M.2 slot for a second SSD but, honestly, that would be a case of "who are you and what have you done with the real Tim Cook?".

I think that "user upgradeable RAM" in Macs now has little to do with the user and everything to do with Apple's own costs and logistics: is type x RAM cheaper in loose SMT or DIMM form? Can we use the same chips that we buy by the shedload at huge bulk discount for the MacBook Air? How many units of the highest RAM capacity will we sell and is it economic to make that number of main boards with fixed RAM or do we need the ability to add the RAM after manufacture? The decision will be made by a spreadsheet and it will be the best, most courageous decision ever.

 

[Joe The Dragon]

...which is a bit of a cheek considering that the desktop chips they are using are considerably cheaper than the successors to the mobile processors used in previous Minis. OK, you get good CPU performance, but the downside is that Intel don't have a "desktop" equivalent of their premium Iris/Pro/Plus/whatever integrated graphics, so the Mini is stuck with lowest-common-denominator graphics (...please don't suggest that I get a neat, tiny computer like a Mini and then tether it to a honking great eGPU chassis that adds at least $200 to the cost of a limited choice of supported GPUs, needs its own wall-wart etc.) - I'd have thought that - circa 2018 - something like the Skull Canyon NUC chipset or those Intel CPUs with in-package discrete AMD GPUs would have produced a better balance between CPU and GPU.

It sounds like Apple Silicon is going to have better integrated graphics than any Intel processor - potentially solving that problem - but it remains to be seen whether the first ASi-for-Mac chip is going to be able to compete with desktop Intel processors on the CPU front (we know that even the A12Z can give a mobile i7 a run for its money). Apple will have to cross that bridge at some stage if they're going to replace the 5k iMac and Mac Pro but I think it's a fairly safe bet that the first ASi chips will be targeted at the lower-end laptops, which should be the easiest win for ASi.

The current Mini design is definitely ripe for an update, seeing as it is still based around a case originally designed to house an optical drive and/or various combinations of spinning rust. So it has an over-abundance of internal space: the 2018 Mini used this to fit in a desktop-grade heatsink and fan. That's probably why it was chosen for the DTK too - room for a slightly kludgey small-run (for Apple) logic board, and the closest thing to a simple headless desktop Apple do (...the DTKs are going to be rounded up and shredded after a year, no point in wasting a whole laptop). So I wouldn't take the DTK as a sign of some new enthusiasm for the Mini on the part of Apple.

Personally, an ASi Mini would be my preferred choice for a first ASi Mac - partly because it would have to work alongside my existing iMac for the near future until everything I need got ported and, longer term, I really don't want another all-in-one. However the reality is that the Mini has never been a big priority to Apple and I'm really not holding my breath.

If they do upgrade the Mini - fixed SSD sucks but is probably inevitable (really don't see Apple back tracking on that) and, in any case, it's going to be proprietary ASi-driven SSD. They could add a standard M.2 slot for a second SSD but, honestly, that would be a case of "who are you and what have you done with the real Tim Cook?".

I think that "user upgradeable RAM" in Macs now has little to do with the user and everything to do with Apple's own costs and logistics: is type x RAM cheaper in loose SMT or DIMM form? Can we use the same chips that we buy by the shedload at huge bulk discount for the MacBook Air? How many units of the highest RAM capacity will we sell and is it economic to make that number of main boards with fixed RAM or do we need the ability to add the RAM after manufacture? The decision will be made by a spreadsheet and it will be the best, most courageous decision ever.

they can use the same ssd slots that mac pro / imac uses in the mini to give storage flex and to cut down on the number of boards with no ram slots.

 

[gkarris]

Just found this.

Apple Silicon Mac mini could just be mightier... and mini-er

Just how much mightier, how much mini-er, could Apple Silicon really make the Mac mini?

www.imore.com

Maybe the ATV will be the new Mac Mini... (all it needs is a change of ports in back)...

 

[sanfrancisofont1984]

I thought the CPU horsepower of AS is already there? Am I reading https://browser.geekbench.com/ios-benchmarks correctly? 1500+ single core score that no x86 chip has archived on average?

 

[T'hain Esh Kelch]

I presume it will be even more mini due to the lack of heat from the new chips

Maybe it'll be appletv size ?

Highly unlikely that they will launch a Mac Mini without active cooling.

 

[ADGrant]

I presume it will be even more mini due to the lack of heat from the new chips

Maybe it'll be appletv size ?

I don't think so. There are a lot of server racks build for the current Mini Chassis and the size allows for more ports on the rear and a decent cooling solution. Could be active but a large passive heat sink would be an interesting idea.

 

[Erehy Dobon]

Forget about user replaceable m.2 NVMe boot SSDs in the ASi Mac mini. It isn't going to happen.

The T2 Security Chip in the Mac mini 2018 is the proprietary disk controller. By using their own silicon, Apple can enforce greater security (tamper protection) versus going to a third-party SSD disk controller chip like Phison, SiliconPower, etc.

Following the evolution of iDevices and Macs, Apple is clearly moving to their own proprietary silicon for many of these tasks.

I can see an m.2 slot as an auxiliary storage option for some of the larger Macs.

Shipping a Mac mini without active cooling makes zero sense. It's a desktop computer powered by AC, not battery power. Apple has more flexibility with thermal design on the Mac mini than their notebook computers. That's why the 2018 has a desktop CPU.

Note that Apple's design philosophy for their desktops has long meant higher operating temperatures than similar Windows computers to squeeze out more performance.

This goes back to primary point from Johny Sroudji's presentation during WWDC keynote: performance per watt. The current Mac mini chassis design has a maximum TDP and I expect Apple to use as much of it as possible.

I would love to see an ASi Mac mini in a similar form factor to the AirPort Extreme employing an 80mm or 92mm fan in the base but I know it will never happen. I would also love to see them remove the power supply from the case but that will never happen either.

 

[theSeb]

I thought the CPU horsepower of AS is already there? Am I reading https://browser.geekbench.com/ios-benchmarks correctly? 1500+ single core score that no x86 chip has archived on average?

Synthetic benchmarks are mostly good for one thing only: to compare which hardware is faster at running that benchmark

 

[Erehy Dobon]

Synthetic benchmarks are indeed pretty stupid. Geekbench is so retarded that the Windows PC hardware reviewers don't even include it (AnandTech, Tom's Hardware, etc.).

Geekbench is all that Mac users can talk about because there is.

Real Windows PC reviews lean heavily to actual game performance (often running 8-14 different games) and more real-world tests like Handbrake encodes or Blender tests. If a PC review is 16 pages, maybe one page covers synthetic benchmarks and 12 pages will cover game performance (one game per page).

The main interest in Geekbench is that a given chip exists. A new chip showing up on the Geekbench site is merits attention but the actual score does not. There's really no telling who is running the benchmark under what circumstances.

 

[sanfrancisofont1984]

Oh I'm sure these iPhone scores are obtained with a huge AIO liquid cooling block attached /s

https://www.tomsguide.com/us/new-ipad-pro-benchmarks,news-28453.html no gaming for sure but PC gaming on Mac is not viable since Catalina already.

 

[theSeb]

I am not sure if you have ever used Geekbench, but when you do you may notice that it's doing a lot of cryptography tests. This skews the results because iDevices have hardware acceleration for many of these specific operations, whilst x86 does not. Back in the day one of the previous versions did not even stress a Mac CPU to cause throttling, so mobile Mac scores were biased against desktop Macs. Using one synthetic benchmark as a data point and trying to extrapolate that to real world applications is generally not a good idea.

Regarding the other benchmarks you've linked, this really shows the benefits of running on a cut down and super optimised operating system vs having to deal with the bloat of a "full" operating system such as Mac OS or Windows and the advantages of close cooperation between the software and the hardware.

Having said that there is no denying that Intel has been sleeping for a decade allowing AMD to catch up and become a player again. Meanwhile Apple has been super busy in the background creating some amazing chip designs. The fact is that since they announced it along with pretty concrete dates for the transition means that everything is planned and pretty much known to them. They are clearly super confident. You don't run a trillion dollar company by steering blindly into the fog and hope for the best and they aren't going to start doing that now.

 

[sanfrancisofont1984]

Crypto is like 1 out of 3 areas http://support.primatelabs.com/kb/geekbench/interpreting-geekbench-5-scores I don't use GB very often as the free version uploads system info. I usually use Cinebench + Unigine Heaven to sanity check my computer builds.

 

[theSeb]

Yes, that sounds about right, so 1/3 of the score is biased towards a platform

 

[Erehy Dobon]

I usually use Cinebench + Unigine Heaven to sanity check my computer builds.

I have quickly learned that Unigine Heaven is great for GPU fan curve tuning. A few months ago I built my first custom Windows PC in since the Nineties and it's interesting to compare the environment in the Nineties versus today.

I moved almost all of my components from my NZXT H1 case to a micro-ATX case. In doing so I also moved my CPU to a 240mm AIO cooler and modded my graphics card to accommodate a 120mm AIO cooler.

Today the graphics card (Nvidia GeForce RTX 2070 SUPER) maxxes out at 65 deg Celsius at 100% GPU load during a stress test on MSI Kombustor. Unigine Heaven is about the same.

I generally discount synthetic benchmarks but games often push the GPU to 99% consistently. I've ran some games' built-in benchmarks (Shadow of the Tomb Raider, Rainbow Six Siege, Horizon Zero Dawn) and they all end up the same way.

Real gameplay is closely in line with this.

I've ran the CPU-Z stress test as well as real-world Handbrake encodes. With careful observation and thoughtful analysis, it's not so difficult to get great performance out of a custom Windows PC build if sane decisions were made in component decision.

 

[sanfrancisofont1984]

Not sure if Handbrake encode is very real world. It could be a cross platform benchmark nonetheless. I use Linux on non-Mac machines.

 

[thingstoponder]

Wouldn't it be crazy if the iMac Pro was in the first batch of Apple Silicon Macs? Obviously the current gen is long-in-the-tooth. Not sure if an overclocked A14Z would bench better than the i9 you can get in the 2020 iMacs.

Probably not because it will only have 4 big cores, but I n single core it undoubtedly would. Whatever Apple Silicon chip that will replace the iMac Pro would need ECC memory support and would need to range in core counts from ~8-18 because I doubt they’re having less cores than the previous model. This is why the iMac and Mac Pro will almost certainly come much later. The 5nm process will need to mature to support these large chips. Coming out the gate with huge chips doesn’t seem to make sense strategically or even seem feasible.

That said this thread seems still be under the assumption that Apple is going to just use iPad chips in Macs. This has no basis given what they’ve said which is that there’s going to be a family of Mac chips and no credible rumors/leaks saying they will. The Mac mini will probably reuse MacBook or iMac chips, not iPad chips.

 

[thingstoponder]

I thought the CPU horsepower of AS is already there? Am I reading https://browser.geekbench.com/ios-benchmarks correctly? 1500+ single core score that no x86 chip has archived on average?

Correct.

Synthetic benchmarks are mostly good for one thing only: to compare which hardware is faster at running that benchmark

Nope. They’re indicivative of real world performance considering they do real world tasks. Geekbench 5 is a great benchmark and results are absolutely valid cross platform.

Synthetic benchmarks are indeed pretty stupid. Geekbench is so retarded that the Windows PC hardware reviewers don't even include it (AnandTech, Tom's Hardware, etc.).

Geekbench is all that Mac users can talk about because there is.

Real Windows PC reviews lean heavily to actual game performance (often running 8-14 different games) and more real-world tests like Handbrake encodes or Blender tests. If a PC review is 16 pages, maybe one page covers synthetic benchmarks and 12 pages will cover game performance (one game per page).

The main interest in Geekbench is that a given chip exists. A new chip showing up on the Geekbench site is merits attention but the actual score does not. There's really no telling who is running the benchmark under what circumstances.

Plenty of Mac reviews use real world tests and plenty of PC reviews use geekbench, not sure what you’re talking about. Apple users probably use it so much because you can compare x86 Macs to Arm iOS devices and that is useful because everyone has known for years that they’ll eventually use their own chips for Macs so people liked to keep tabs of how the performance compared.

 

[theSeb]

Nope. They’re indicivative of real world performance considering they do real world tasks. Geekbench 5 is a great benchmark and results are absolutely valid cross platform.

Please tell me more with some examples of the real world tasks.