DigiTimes: Apple to Announce Redesigned 14-inch and 16-inch MacBook Pro in September

[transpo1]

Actually it might make sense to release the larger M2X before the smaller M2. The production volumes of the larger chip are smaller, so there is less time pressure to produce a large quantity, and any delays or issues in fabrication would have less of an impact of overall production.

I think @cmaier also mentioned that there are no technical reasons why a larger chip can't precede a smaller chip.

I was thinking this, too. Releasing the more powerful chip first and letting it create a "halo" around the pro products before letting it trickle down to the consumer laptops. But this is problematic, too, as it would be confusing to release an M2X before introducing the M2.

I think the only way for them to go here is to introduce "M2" and then later an "M2X." Or introduce the M2 for the Pro products first and then later use it for next year's base model laptops, followed by the M3 for the following year's Pros and so on...

 

[dwmc]

I’m in the film industry so I understand exactly what needs Thunderbolt and what doesn’t.

Please explain to us your use case instead of whining and maybe the good folks on here can provide you some solutions. Some have responded already.

There are many use cases where the accessible Thunderbolt bandwidth is important. Bandwidth saves time.

I'd miss a 4th Thunderbolt port more than I've missed MagSafe (despite the fact that I loved MagSafe in its day). The magnetic Thunderbolt adapters I bought have been collecting dust in a drawer for a while now... they work, but for me they're mostly novelty. I have no need for HDMI or an SD reader, but I'd not turn down CFExpress. But my needs are my needs, not everyone's needs. It's a laptop, there will be compromises.

For some of us, the availability of high bandwidth, flexible I/O for the next MBP is a source of anxiety. That's not without reason, given what we've seen from M1. A wait and see situation, and not unfamiliar.

 

[transpo1]

There are many use cases where the accessible Thunderbolt bandwidth is important. Bandwidth saves time.

I'd miss a 4th Thunderbolt port more than I've missed MagSafe (despite the fact that I loved MagSafe in its day). The magnetic Thunderbolt adapters I bought have been collecting dust in a drawer for a while now... they work, but for me they're mostly novelty. I have no need for HDMI or an SD reader, but I'd not turn down CFExpress. But my needs are my needs, not everyone's needs. It's a laptop, there will be compromises.

For some of us, the availability of high bandwidth, flexible I/O for the next MBP is a source of anxiety. That's not without reason, given what we've seen from M1. A wait and see situation, and not unfamiliar.

I can see how you'd miss a 4th TB port, as I've needed to DIT from multiple card readers to multiple drives on a film set (requires 4 ports) from time-to-time but this can be achieved with certain USB-C hubs that deliver power. I'd gladly sacrifice a 4th TB port for not having to worry about whether I have an SD card dongle or HDMI with me.

I'm curious- what is your particular use case?

 

[aapl owner]

Trying to limit damage to sales of current M1 devices by keeping very quiet.

Huh, Apple will announce the new MacBook Pro's when they are ready, who in their right mind would do otherwise.

 

[Zdigital2015]

Huh, Apple will announce the new MacBook Pro's when they are ready, who in their right mind would do otherwise.

That why I highly doubt anyone on these forums works for Apple. Having worked with Apple as client, most would never make it to the interview process. The mindset is just not there. That’s not meant as an overbroad statement, that are some wonderful commenters here.

 

[Flood4]

This isn’t a networking forum, so it isn’t like there would be cause for confusion. Anyone reading the post would understand the poster is discussing a computer, not the physical address burned into a networking chip, or a randomized address derived in software like iOS does with public Wi-Fi hotspots.

True. That’s why I mentioned it was a FYI.

 

[Minxy]

So the later release means this is more likely going to have the M2 chip?

 

[sublunar]

This is not likely at all. We aren't "skipping" anything. The chip that is in the new MacBook Pro existed a year ago, and was reported on. They didn't just move to a new chip architecture because it has taken a couple extra months to get it out the door. The product was designed from the start with the M1X.

It is also going to be announced and shipped several months before any "M2" product. You think they're going to introduce the M2X months before the M2? Lol.

Let’s bear in mind that the A14 chip debuted in the iPad Air 4 which was announced before even the (delayed) iPhone 12 came out - leading people to be able to better predict benchmarks ahead of the actual phone release based on Apple’s earlier presentation.

With ARM CPUs becoming the way ahead there increasingly doesn’t need to be any delay between events, there’s so many reasons why Apple could choose to announce a product at a specific time.

But by the same logic, it doesn’t make sense for Apple’s most expensive MBPs to be behind the MBA when it comes to silicon. The “X” versions will barely be out the door when the next-gen silicon will be showing up in the MBA, etc. That seems less than optimal from a marketing standpoint and confusing for the average consumer.

Let’s just see how Apple intend to construct their CPU series. Annual CPU updates occur because of the phones - it’s up to Apple how they package them for Macs, iPads or iPhones.

It’s easy to assume that iPhones will be 2+4 core annual CPU updates because that’s the way it’s been for years.

Apple may choose to create a new M1-style 4+4 CPU every other year for the performance jump (based on the iPhone cores of that year), adding additional cores in the between years during a process shrink so the power envelope remains consistent and that there is also an ‘annual’ update for iPad/Mac users.

It would be Apple’s version of the much vaunted Intel tick-tock cadence. The M1 is on TSMC’s 5nm process, this year’s A15 is apparently on an enhanced 5nm process while next year’s A16 could be on a 4nm process.

The heavy duty CPUs expected in high end Macs - the 14”, 16”, iMac Pro, Mac Pro could be on a similar routine - only with immensely more cores and higher power envelope.

On this basis I would estimate that A16 in 2022 will be the chip family to look out for in the high end desktops - it fits quite nicely into the idea that the Mac Pro and maybe iMac Pro will be coming along in (for example) October next year.

This year will bring enhancements to the M1 to allow extra cores for the same power envelope - perfect for the 14” and 16” MBP - possibly even the Mac Mini. If this arrangement is ok for a 27” iMac replacement this year Apple could bring that online with a more powerful GPU unless they are looking to create a more iMac Pro replacement - and we won’t see that till next year.

Where does this leave the existing M1 products? Do people expect the M1 Macs to get the M1X CPU? It depends if the per core performance is a significant improvement on the M1 but another factor is the staggered release schedule - the 24” iMac only just got the M1 so would it get the M1X so soon?

 

[dwmc]

I can see how you'd miss a 4th TB port, as I've needed to DIT from multiple card readers to multiple drives on a film set (requires 4 ports) from time-to-time but this can be achieved with certain USB-C hubs that deliver power. I'd gladly sacrifice a 4th TB port for not having to worry about whether I have an SD card dongle or HDMI with me.

I'm curious- what is your particular use case?

Broad spectrum of software development (from 8-bit micros to Epyc servers). Not long ago I was doing ML work. Hence a lot of shuttling of large data sets from the field to the home office, for which the MBP has been effective. When docked at home, the MBP sits on two 10G ethernet networks (both private, one is dedicated to SAN) via two CalDigit Connect adapters, and a 1G network (for route to gateway to public Internet). Sometimes has eGPU plugged in (for GPGPU, not display). Sometimes has Thunderbolt storage plugged in (for larger sets of training data and scratch space). 1G network, external display, power and other stuff via one of the three TS3+ I own (good docks, I've had them for years and they just work). Most of the time, my MBP is in clamshell mode at home, I use it remotely via ssh and screen sharing from my desktop or flip the KVM built into my desktop monitor and swap the magic trackpad with the one that's paired to the MBP.

I've spent much the last 30 years working from my home office. I have robust unattended backups for everything (over the network to ZFS pools dedicated to backups). That's the motivation for one of the 10G networks (SAN). I have OM4 and OM5 fiber in the walls, as well as Cat6A to every room that needs it. I'd go to 40G if it made financial sense. The fiber is cheap (cheaper than Cat6A), it's all the other stuff that gets expensive but eventually it'll be in reach. NVMe has raced ahead for years now, and for me, making effective use of it across the board means my network has to keep up. I'm assuming the next MBP will have internal storage that can easily fill a 40Gbits/sec pipe, if not two. I could make good use of that if it weren't for the price of the switches, NICs and additional NVMe I'd need to put in place.

I can carry what I need. Big backpack, and nearly everything in it is dedicated for field use (pretty much the only things that ever come out at home are the MBP and portable storage drives). In other words, I never use any dongle, cable or adapter from my bag at home. If I need one at home, I buy one for home. And maybe a third as a spare. In the bag, I have multiples of some things and a pair of small USB-C multi-purpose adapters. What I can't do is add a Thunderbolt port to the MBP where one doesn't exist.

The MBP has kept up with the network updates, happily living on the three networks at once while still being connected to eGPU and Thunderbolt storage and the 'other stuff'. Gracefully. I'm honestly more anxious about overall bandwidth in the next MBP 16" than I am the number of ports, as long as it's at least three. But if it's three with fine print that explains the 3 ports are split across the equivalent of 2 PCIe Gen4 lanes...

 

[Fomalhaut]

So the later release means this is more likely going to have the M2 chip?

I would say so, yes. At least that's what I'm hoping. At the beginning of a new product's lifespan, you would hope for rapid innovation to include all the features that couldn't be included in the previous version due to limitations of time.

2nd generation products are often a big improvement over the first one, and after that the level of improvements in each version usually diminish.

 

[Freedom1]

After the WWDC debacle, it seems foolish to believe ANY leaks. They all seem click-bait. Anyone who turns out “right” was almost surely just edu-guessing.

 

[transpo1]

Someone's always going to be disappointed. But you can always get a USB-C hub that replaces the need for one more USB-C port quite easily. And for most people, having one of the other slots (HDMI or SD card slot) replaces the need for the 4th port anyway.

I should note, that the best case scenario would be 4 USB-C ports, MagSafe, an SD card slot, and HDMI for those who need it. But that won't happen

 

[Stephen.R]

But you can always get a USB-C hub that replaces the need for one more USB-C port quite easily.

Which bus powered hub are you referring to that can provide two downstream TB3/USB4 ports from a single upstream port?

 

[jk73]

Broad spectrum of software development (from 8-bit micros to Epyc servers). Not long ago I was doing ML work. Hence a lot of shuttling of large data sets from the field to the home office, for which the MBP has been effective. When docked at home, the MBP sits on two 10G ethernet networks (both private, one is dedicated to SAN) via two CalDigit Connect adapters, and a 1G network (for route to gateway to public Internet). Sometimes has eGPU plugged in (for GPGPU, not display). Sometimes has Thunderbolt storage plugged in (for larger sets of training data and scratch space). 1G network, external display, power and other stuff via one of the three TS3+ I own (good docks, I've had them for years and they just work). Most of the time, my MBP is in clamshell mode at home, I use it remotely via ssh and screen sharing from my desktop or flip the KVM built into my desktop monitor and swap the magic trackpad with the one that's paired to the MBP.

I've spent much the last 30 years working from my home office. I have robust unattended backups for everything (over the network to ZFS pools dedicated to backups). That's the motivation for one of the 10G networks (SAN). I have OM4 and OM5 fiber in the walls, as well as Cat6A to every room that needs it. I'd go to 40G if it made financial sense. The fiber is cheap (cheaper than Cat6A), it's all the other stuff that gets expensive but eventually it'll be in reach. NVMe has raced ahead for years now, and for me, making effective use of it across the board means my network has to keep up. I'm assuming the next MBP will have internal storage that can easily fill a 40Gbits/sec pipe, if not two. I could make good use of that if it weren't for the price of the switches, NICs and additional NVMe I'd need to put in place.

I can carry what I need. Big backpack, and nearly everything in it is dedicated for field use (pretty much the only things that ever come out at home are the MBP and portable storage drives). In other words, I never use any dongle, cable or adapter from my bag at home. If I need one at home, I buy one for home. And maybe a third as a spare. In the bag, I have multiples of some things and a pair of small USB-C multi-purpose adapters. What I can't do is add a Thunderbolt port to the MBP where one doesn't exist.

The MBP has kept up with the network updates, happily living on the three networks at once while still being connected to eGPU and Thunderbolt storage and the 'other stuff'. Gracefully. I'm honestly more anxious about overall bandwidth in the next MBP 16" than I am the number of ports, as long as it's at least three. But if it's three with fine print that explains the 3 ports are split across the equivalent of 2 PCIe Gen4 lanes...

Reading this brings Niagara Falls to mind. That’s a lot of data moving around. Wow.

 

[Stephen.R]

I should note, that the best case scenario would be 4 USB-C ports, MagSafe, an SD card slot, and HDMI for those who need it. But that won't happen

That very much depends on the implementation though.

In the case of a hypothetical MBP with say 4 TB3/USB4 ports, MagSafe, SD slot and HDMI - if the HDMI port has a hardwired video stream from the GPU (e.g. as the Mini has had since 2018) - that's worse than the predecessor, unless those GPUs somehow support six displays (built-in, 4xTB3, HDMI). At this point there's no reason necessarily to assume the machines will even support as many displays as their predecessors.

 

[transpo1]

Which bus powered hub are you referring to that can provide two downstream TB3/USB4 ports from a single upstream port?

USB-C bus-powered hubs exist. I doubt TB3 bus-powered hubs do. (Perhaps selfishly I was thinking about my own use case wherein I would sometimes need to have two card readers / two USB-C SSDs plugged in at once.)

The OWC Thunderbolt Hub is not bus-powered, but offers three TB3 from one upstream.

 

[Stephen.R]

USB-C bus-powered hubs exist.

I know. And USB-C to HDMI adapters/cables also don't require AC power. Neither do most USB-C or even TB3 card readers.

The OWC Thunderbolt Hub is not bus-powered, but offers three TB3 from one upstream.

I know that too.


That's kind of why I replied. Suggesting that a USB3.x type-c hub is a replacement for a fourth tb3/usb4 port, is a huge stretch/assumption about what the person needs the fourth port for.

 

[erich.j.k]

These were shipping almost immediately last month. If true, either BTS sales are strong or Apple is gearing up to release new MBPs.

Seems they still are shipping immediately (non-bto 16").

 

[erich.j.k]

I was thinking this, too. Releasing the more powerful chip first and letting it create a "halo" around the pro products before letting it trickle down to the consumer laptops. But this is problematic, too, as it would be confusing to release an M2X before introducing the M2.

I think the only way for them to go here is to introduce "M2" and then later an "M2X." Or introduce the M2 for the Pro products first and then later use it for next year's base model laptops, followed by the M3 for the following year's Pros and so on...

I believe Apple originally wanted to introduce the new MacBook Pros with a M1X by now. And then something went wrong.

I remember an article about Samsung 16GB modules that Apple would need for the SoC being too expensive (because the production is still slow). Then there was the mini-LED shortage. Maybe it was something else altogether.

Now that they can't show us the MacBook Pros (or larger iMacs or Mac minis), they're totally thrown off. I doubt the M2X is anywhere near ready (even though I'd be very glad to be wrong on this). So they have a choice of using an outdated M1X around October (with a M2X planned for spring 2022) or a M2 - which is probably what should have been the successor for the M1 in the lower end.

So, now what? Wait til the M2X is ready? That will probably be spring 2022 - earliest. A long wait for me...

 

[jk73]

So, now what? Wait til the M2X is ready? That will probably be spring 2022 - earliest. A long wait for me...

Hard to imagine a two-month delay (from July to September) would then cause a 9-month delay.

Apple would probably just rename the M1X the M2 and then sort everything out later.

It sounds like Apple could’ve had these out the door already if not for the screen supply.

 

[erich.j.k]

Hard to imagine a two-month delay (from July to September) would then cause a 9-month delay.

IMHO July would have been the M1X. Spring (or maybe even summer) 2022 would be the M2X.

Makes me wonder if I should wait for the second generation "pro"-SoCs. Probably won't though since I'd really like to check the Mx out but want more RAM than they currently offer in the M1 SoCs.

 

[jk73]

IMHO July would have been the M1X. Spring (or maybe even summer) 2022 would be the M2X.

Right, but I don’t see any chance of Apple deciding that because July turned into September, they might as well just wait until spring 2022. If nothing else, what would Apple do with all of the 14-inch and 16-inch screens that are supposedly starting to arrive this month? I could see two MBP updates in 8 months (M1X in Sept., M2X in 2022) before I could see no 16-inch MBP update for 30 months.

 

[transpo1]

I know. And USB-C to HDMI adapters/cables also don't require AC power. Neither do most USB-C or even TB3 card readers.


I know that too.


That's kind of why I replied. Suggesting that a USB3.x type-c hub is a replacement for a fourth tb3/usb4 port, is a huge stretch/assumption about what the person needs the fourth port for.

I simply said that there are USB-C hubs that can help. For my use case, with the concern of being without a dongle for simple things like having an SD card reader when you need one, it’s worth it for me to have those built in.

If Apple is truly going to release a MacBook Pro with an SD card reader and HDMI again, they must have done their research and found that the majority of users miss having these ports around. YMMV.

 

[sublunar]

I simply said that there are USB-C hubs that can help. For my use case, with the concern of being without a dongle for simple things like having an SD card reader when you need one, it’s worth it for me to have those built in.

If Apple is truly going to release a MacBook Pro with an SD card reader and HDMI again, they must have done their research and found that the majority of users miss having these ports around. YMMV.

Apple have been working back some of their previous choices from recent years so after reverting the magic keyboard and potentially getting rid of the Touch Bar are major design decisions. I shouldn’t have thought that swapping 1 out of 4 USB-C/Thunderbolt ports to bring back HDMI and SD card slots would be a bad design either. And don’t forget a MagSafe port brings back a third port too - it’s entirely feasible that Macbook Pro users on the road would have used up a Thunderbolt port with a USB-C charger.

Even with one port taken up by a charger, how many of us on the road really use all 4 TB ports at the same time?

Plus, it‘s also potentially a cost saver - allowing Apple to go with 3 TB controllers instead of 4 (even with a licensing fee the HDMI port and SD port should be cheaper to implement). And if, at a pinch, a generic USB-C charger can still be used to charge from one of the 3 USB-C ports then the flexibility to take cheaper chargers on the road would be a good move.

 

[transpo1]

Apple have been working back some of their previous choices from recent years so after reverting the magic keyboard and potentially getting rid of the Touch Bar are major design decisions. I shouldn’t have thought that swapping 1 out of 4 USB-C/Thunderbolt ports to bring back HDMI and SD card slots would be a bad design either. And don’t forget a MagSafe port brings back a third port too - it’s entirely feasible that Macbook Pro users on the road would have used up a Thunderbolt port with a USB-C charger.

Even with one port taken up by a charger, how many of us on the road really use all 4 TB ports at the same time?

Plus, it‘s also potentially a cost saver - allowing Apple to go with 3 TB controllers instead of 4 (even with a licensing fee the HDMI port and SD port should be cheaper to implement). And if, at a pinch, a generic USB-C charger can still be used to charge from one of the 3 USB-C ports then the flexibility to take cheaper chargers on the road would be a good move.

Exactly. 💯