MacBook Pro Now Expected to Get M5 Chip Before iPad Pro

[MacRumors]

Apple will bring its next-generation M5 chip to the MacBook Pro in the fall, followed by the iPad Pro in the first half of 2026, according to Bloomberg's Mark Gurman.

Gurman's latest claim flips the previously expected order of devices to transition from M4 to M5 chips, which was based on last year's M4 product launch order. Apple first introduced its M4 chip in an updated iPad Pro in May 2024, followed by the MacBook Pro in October, but it sounds like Apple will deviate from that timeline this time around.

Before the release of M5 MacBook Pro models, Gurman says that Apple will launch updates to the Mac Studio and Mac Pro using its current generation M4 chip series. Those machines could arrive around Apple's annual Worldwide Developers Conference in June 2025.

The M5 series is expected to feature an enhanced ARM architecture and is reportedly being manufactured using TSMC's advanced 3-nanometer process technology. Apple's decision to forgo TSMC's more advanced 2nm process for the M5 chip is believed to be due to cost considerations. However, the high-end versions of the M5 will still feature significant advancements over their M4 equivalents, mainly through the adoption of TSMC's System on Integrated Chip (SoIC) technology.

This 3D chip-stacking approach vertically stacks the chips, which enhances thermal management and reduces electrical leakage compared to traditional 2D designs. Apple is said to have expanded its cooperation with TSMC on the next-generation hybrid SoIC package, which also combines thermoplastic carbon fiber composite molding technology.

References to what are believed to be Apple's M5 chip have already been discovered in official Apple code. According to one report, thanks to its dual-use SoIC design, Apple also plans to deploy the M5 chip within its AI server infrastructure to bolster AI capabilities across both consumer devices and cloud services.

Article Link: MacBook Pro Now Expected to Get M5 Chip Before iPad Pro

 

[BelgianChoklit]

And imo that’s how it should be. The MacBook Pro is a machine that can actually put the power to good use.

 

[BigDO]

The M4 iPad Pro is about 10 years ahead of what iPadOS can do with it anyway.

 

[Piggie]

Can anyone update me please?

Is the M5 supposed to be an interesting hike up in tech from the M4 with key differences?
Or is it more of one of the other M series changes, where it's pretty much a light polish and tweak of the previous revision?

 

[Kylo83]

It’s all about the m6 with the total redesign

 

[throAU]

Can anyone update me please?

Is the M5 supposed to be an interesting hike up in tech from the M4 with key differences?
Or is it more of one of the other M series changes, where it's pretty much a light polish and tweak of the previous revision?

It will be an evolution.

I wouldn’t call previous m series upgrades a light polish. They’re getting much larger gains than intel and amd in the same time period.

Which is impressive. The low hanging fruit was taken decades ago.

50-100 percent extra performance in 3 years (eg M4 Pro beating M1 Max with more cores, and measured +50 percent in single thread) is impressive. Intel has been coasting on +5 percent per year for decades.

 

[wanha]

Can anyone update me please?

Is the M5 supposed to be an interesting hike up in tech from the M4 with key differences?
Or is it more of one of the other M series changes, where it's pretty much a light polish and tweak of the previous revision?

While we don't have any concrete, provable facts on this, rumors point to the M4 still using TSMC’s third-generation N3P process, which should translate to modest performance and efficiency improvements over the M4.

The M6 is rumored to be based on the more advanced 2nm process, which is likely to translate into more significant performance and efficiency gains.

That said, it's good to keep in mind that the M5 will be THE state of the art chip in consumer laptops in 2025 and I don't think anyone will regret buying one.

 

[iBluetooth]

Can anyone update me please?

Is the M5 supposed to be an interesting hike up in tech from the M4 with key differences?
Or is it more of one of the other M series changes, where it's pretty much a light polish and tweak of the previous revision?

With chiplets they are likely to pack together based on the focus of the device, like increasing the GPU or NPU (AI), or more focus on compute with added cores. They got caught behind with the AI, which will take them a time to catch up (chip from initial design to production is 4 ~ years).

 

[Cervisia]

50-100 percent extra performance in 3 years (eg M4 Pro beating M1 Max with more cores, and measured +50 percent in single thread) is impressive. Intel has been coasting on +5 percent per year for decades.

The M4 Pro has more cores...

 

[iToph]

Dear admins, can you please rename to: Gurmanrumors.com

 

[diggy33]

As it should be.

 

[G3Blue]

,

 

[Nismo73]

It might get it before the MacBook Air gets the M4.

 

[throAU]

The M4 Pro has more cores...

Right you are - I forgot the M1 Pro/max had same CPU core count. It doesn't have more GPU cores though and that's also faster. Single core is also in the order of +50% (per core).

Main point still stands, these aren't minor incremental improvements over 3 years, in 3 generations there have been real, meaningful upgrades to responsiveness and outright throughput.

Don't get me wrong - not saying "M1 is crap now!" or anything like that. But claiming there's only been minor improvements to performance... anyone saying that just hasn't been paying attention (or used the products back to back). Even A14 generation cores to A15 (i.e., M1 to M2) was a real leap in single thread interactive responsiveness.

 

[AndreDC]

Dear admins, can you please rename to: Gurmanrumors.com

Domain available.... you know you want to....

 

[MRSugarD]

The M5 should be the last chip update before the MBP design overhaul, no?

 

[svish]

Possible. But since there are rumors that M5 is already under production, it is possible for it to get a M5 update at this year's WWDC. Current iPad Pro is more than enough for any task and a processor upgrade is not really required.

 

[leehericks]

Mac Studio users deserve an M5 Ultra update this year, not to be held back with an M4…just postulating.

 

[Unami]

Makes sense to first release it in a product that has the software and OS to actually make use of it.

 

[Piggie]

What I'd love (and I'm sure millions of others also would) would be for Apple to tackle the GPU side of things.
When it comes to, let's a say an M4 Mini, it's an amazing machine, CPU is stunning, but the GPU, whilst of for what it is, still needs massive work to make this into the machine is could be.

Yes, we all know, due totally to Steve Jobs and years of pettiness between Apple/Nvidia, Apple killed itself on the GPU front. But perhaps, (If they have the willpower, which is the key) they can now change this.

Part of me wishes Apple would create its own separate GPU which could be fitted alongside an M Series chip for those who want it.
Unless of course it ever becomes possible to enhance the on-chip GPU to a vastly higher level.

Apple COULD make the Mac Mini a best selling little gaming powerhouse, but sadly part of me still feels in 10+ years time we'll still be waiting for them.

Hope I'm wrong.

 

[HobeSoundDarryl]

Mac Studio users deserve an M5 Ultra update this year, not to be held back with an M4…just postulating.

I actually think this would make sense- to lead with M5 MAX & ULTRA instead of saving ULTRA for last in a line. As an ULTRA owner, I would not do it again, knowing that Mnext MAX is no more than 6 months away and is likely to be about as powerful and cost substantially less. IMO: the (up to) 6 months as "king" is not worth it.

However, flip the apparent release schedule slots of BASE vs. ULTRA and that seemingly fixes everything, including maximizing profit for Apple too. MAX & ULTRA first, PRO & MAX in the usual October slot, BASE the following spring. In this flip, one could confidently pay maximum (revenue & profit) for an ULTRA, knowing they have at least a year (maybe even 1.5 years) of perceiving their Mac is most powerful Mac. Bread & Butter MBpros, etc launch on the very same Fall schedule as they do now, so no effect on them. Lowest price (and presumably lower profit) BASE Macs roll out last.

Among other things, this creates the natural pressure to pay up for a higher tier chip (and thus more profit per unit sold for Apple) vs. waiting around for the "bargain" Macs... which would then be the ones bumping into the impending Mnext release speculation. As is, BASE first facilitates "bigger number" rationale further fueled by cheaper pricing for BASE Macs. In other words, for a few months, cheapest price buyers can have the next generation number and still be on "latest generation" (number) for the longest time while other, pricier Macs only step up to the same number... or linger 2 numbers behind.

Speaking just for myself: I would never buy ULTRA again based upon the reality that Mnext no more than about 6 months away is likely to be about as powerful and perhaps more in select ways. So all the extra profit realized from me with that ULTRA purchase will not be realized again without ULTRA owning the crown for longer than a half year or less. No, that is not a call to delay launches at all- just consider flipping the releases around. And no, that should have little-to-no effect on the all-important Fall releases which could still hit at the exact same time.

Presumably, ULTRA is not able to be built first and BASE must come first, then PRO/MAX and lastly ULTRA. But whether that's technical limitations or just how it's been done so far is the big question. It seems Apple could target ULTRA first and then the rest would be engineering derivatives of it over the next year... instead of starting with BASE and building up to ULTRA over the same period of time.

And yes, I know well the counterpoint about market share to try to make sense of the "as is" schedule. I don't know that market share would change much with this concept either... but average profit per unit sold might go up and that seems all-important to modern Apple. 💰💰💰

For example, if guy who might typically buy MBair can't stand the wait for BASE, perhaps they move sooner to MBpro to get "latest chip power" and thus deliver greater revenue & profit for Apple. Maybe guy who might typically buy a Mini is pulled up to a MAX or even ULTRA Studio vs. waiting until the end of a generation to still buy Mini. Etc.

 

[throAU]

What I'd love (and I'm sure millions of others also would) would be for Apple to tackle the GPU side of things.
When it comes to, let's a say an M4 Mini, it's an amazing machine, CPU is stunning, but the GPU, whilst of for what it is, still needs massive work to make this into the machine is could be.

For the mini I think the current SOCs on offer (base m4 / pro) is appropriate.

Its an ultra small form factor box, which means that it needs to run within a particular thermal envelope.

If you were to try stuff a Max or Ultra spec GPU in there, it would either be (pick 2)

• noisy
• overheating
• throttling

... simply because the enclosure and cooling capacity is designed with the current base/pro chips in mind.

I think more so what apple need to do is tackle the studio and more so, the Mac Pro (i.e., ultra high end) side of things.

The current ultra CPUs in the pro are a bit of a joke. They're great for a studio but they're WAY below what the proper full fat Mac Pro should have.

But the rumours indicate the new "hydra" platform is coming and my bet is that like Ryzen->thread ripper evolution before it, Apple will migrate from simply joining two max dies together via ultra fusion, to a proper control die + associated other dies layout. This will enable them to mix/match CPU, GPU and memory dies as appropriate.

So you'll end up with a central die to coordinate things and multiple (i.e., not just 2 combined max CPU/GPU dies) other chips for GPU/CPU linked together with a high speed interconnect (Apple's version of infinity fabric). e.g., I would expect at least 4x M4 Max, if they're committed to building an actual high end platform. I mean the current pro chassis is good to dissipate about 1KW worth of thermal output, and the current M2 Ultra platform is something like 220-250 watts!

But all that said, as per previous posts on here, I think apple are really struggling for the Mac Pro to justify its existence.

Video editing (as a target market for the Mac Pro) is done. You can do a lot of it on a freaking iPad.

High end rendering/3d work - PC is cheaper.

Unless apple properly reset the bar with something properly insane (like the equivalent of 8x M4 Max linked together in the Pro, not just 2 of them) they're going to struggle to justify something thats "a bit higher end than a studio, but not performance competitive with a high end PC workstation" outside of the really tiny niche of video codec processing.

 

[Unregistered 4U]

Can anyone update me please?

Is the M5 supposed to be an interesting hike up in tech from the M4 with key differences?
Or is it more of one of the other M series changes, where it's pretty much a light polish and tweak of the previous revision?

The result will be a light polish and tweak because Apple no longer chases the performance crown. Even if there are key differences, they will lean towards efficiency/longer battery life. It MAY be worth a purchase for M1 owners or below (or people that don’t currently own a computer and are more accustomed to single point interfaces. Which IS an enormous number of people, FAR more than the number of people that own M4’s.

 

[bradman83]

Can anyone update me please?

Is the M5 supposed to be an interesting hike up in tech from the M4 with key differences?
Or is it more of one of the other M series changes, where it's pretty much a light polish and tweak of the previous revision?

It will be an evolution.

While we don't have any concrete, provable facts on this, rumors point to the M4 still using TSMC’s third-generation N3P process, which should translate to modest performance and efficiency improvements over the M4.

I'm going to disagree with the previous quoters. The M5 has the potential to a massive leap over the M1-M4, particularly for higher end chips. The reason is SoIC. Right now all M-series chips are single, monolithic SoC dies save for the Ultra, which just takes two Maxes and fuses them together. Max dies are about as physically large as you can practically make a single die with an acceptable yield. SoIC allows for the components of the SoC to be fabricated separately and then stacked together without a hit to performance and thermal management. Intel does something similar with its tile architecture and AMD uses an older form of the technology with its 3D cache designs.

SoIC likely won't benefit the base M5 chip but it will unlock the potential for Apple to design higher end chips around modular tiles without having to tape out a specific die design for each variant, meaning it could be used to make beefier Max chips and purpose-built Ultra chips (or even something more performant for the Mac Pro).

 

[throAU]

I'm going to disagree with the previous quoters. The M5 has the potential to a massive leap over the M1-M4, particularly for higher end chips. The reason is SoIC. Right now all M-series chips are single, monolithic SoC dies save for the Ultra, which just takes two Maxes and fuses them together. Max dies are about as physically large as you can practically make a single die with an acceptable yield. SoIC allows for the components of the SoC to be fabricated separately and then stacked together without a hit to performance and thermal management. Intel does something similar with its tile architecture and AMD uses an older form of the technology with its 3D cache designs.

SoIC likely won't benefit the base M5 chip but it will unlock the potential for Apple to design higher end chips around modular tiles without having to tape out a specific die design for each variant, meaning it could be used to make beefier Max chips and purpose-built Ultra chips (or even something more performant for the Mac Pro).

I agree with this, but I'm not sure that is coming with M5 generation, I'm still expecting an M4 ultra (hidra) that does this.

But whether it (Hidra) is called M4 or M5, I 100% agree that they'll be stacking/tiling/combining multiple smaller component dies and linking them together with a switching fabric via a control die; much like AMD are currently doing with Threadripper.

And yes, it's all to do with the size/yield limit in silicon manufacturing. Everybody is facing the same issues, Apple isn't immune to the manufacturing issues; they need to deal with the laws of physics, etc. just like everybody else.

They just managed to get away with ultra fusion (equivalent of Threadripper 1000/2000 series just linking 2 Ryzens together) for a little longer due to less architectural baggage with their much newer design, is my bet. Well, that and basically getting top priority on TSMC's best manufacturing process due to their close partnership.