Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
It is so "telling" when the single-core performance isn't listed. As others have said, many of us are not in the Media-Biz and don't care about "graphics"; we want "crank power".

By not listing the M5 single-core stats, it raises the question of whether it is actually lower than the M4.

Maybe someday an AI "pre-compiler" will look at the code and automatically compile it to utilize multi-cores. For now, the code I write, loops and calculations, wants the fastest single-core it can get.
 
  • Disagree
Reactions: wyrdness
This article reads like it was AI generated and not fact checked. In particular "Support for up to 2TB storage" ... this is an Apple configuration max for the assembled M4 Macbook Pro (non-Pro M4) and not a M4 chip limitation.
 
This article reads like it was AI generated and not fact checked. In particular "Support for up to 2TB storage" ... this is an Apple configuration max for the assembled M4 Macbook Pro (non-Pro M4) and not a M4 chip limitation.
It seems to be what the M4 Mac Mini tops out at as well. I haven't seen any third party's being able to offer more than that. Only the M4 Pro chip and above seems to offer storage above 2TB. If there is an M5 Mini, the option of 4TB internal storage might prove more interesting than incremental improvements in processor performance for those who don't need the extra grunt (and cost) of an AS Pro chip.
 
So the MacBook Pro only has a base chip now? Will the Air also have this base M5 chip? Is it the same chip inside the iPad Pro?

Why do they keep screwing with the naming? Pro should have the Pro chip like it's been for M1-M4, what's wrong with these people.
The M5 Pro and Max versions are coming later. It was just the base model MacBook Pro that was updated. The M4 Pro and M4 Max versions are still on sale and will be updated to the M5 equivalents likely after the New Year with the Air following sometime in the spring.
 
While AI and LLMs get all the headlines these days, the underlying neural acceleration hardware boosts a host of other tasks that utilize machine learning routines. Things like the blurred backgrounds in FaceTime calls or copying text out of a photo in the Photos app all use ML and the neural hardware. Third party developers have access to this hardware so even if something isn't billed as AI (although everything today seems to be billed as AI for marketing purposes) it can still get a boost from the hardware improvements.
This is such an important point. Apple's software does a ton of background machine learning that most people have no idea is happening. This has been true for years. Apple first included a "Neural Engine" in the A11 back in 2017 (iPhone 8 and X). That was to handle various "AI" tasks. Apple's machine learning efforts started well before then, but that was the first step with dedicated hardware.
 
This has been discussed for years and its ignored. Similar to how they only compare the latest iPhones to each other and ignore previous gens until a month or more later, if at all.

I believe its down to the fact that they are using the latest device so don't care about other ones.

Also this article having no real data about what it does, because up to 45% faster includes 0% as a number, makes it all pretty useless anyway. I'm sure the new device is faster, but I'm not sure we can state for a fact what it is until they do real world benchmarks.

QFT.

Funny thing is, actually a bit sad, is that the comments from the readers are far more informative than many articles posted here.

Do I have a youtuber for you...

He does exactly what you are asking for. Compares M series SOC`s performance for lightroom etc.



See MR? A 33 minute video, likely to be a 14 plus page pdf once written out, would be solid journalism. Why can't we get to read an in-depth review like this one?

@aibloop: thank you ever so much, really appreciate it!
 
Good to see the usual speed and performance improvements. However those with M4 need not upgrade to M5. Waiting for Apple's whole lineup to be updated with the new chip. Think future M6 with 2nm should offer significant benefits.
 
  • Like
Reactions: arefbe and mganu
I am getting really tired of Apple playing this game of “here’s our brand new processor! OR…you can buy the super maxxed ultra crazy version of the previous processor at a super maxxed premium price!!!”
Not much different to any other company. At the end of the day you buy what you want or what you can afford. I have a Mini M2 pro, happy with it, can't see me changing for a while
 
Apple needs to stop improving their silicon every year because it's angering too many people.
Is this what not happens everywhere? AMD, Intel and all those that use the arm chip on other devices. You buy a nice router and a few months later they bring out another one with better specs. Video cards for P.cs, a new one comes out often offering this and that, phones are the same.

At the end of the day, it is up to the individual if they want to buy it or not. I know buying a mac is different from updating a PC, but even then, with a lot of CPUs, some people may have to buy a new board and memory, no doubt memory.
Oh yeah TV sets, new ones all the time, offering better picture and this and that.

If what you have is doing the job, then no need to change and for personal use most of the time there is no need to change every couple of years or so. My phone is over 4 years old, I am only going to change it because it has a problem, not because of a new shiny phone with better this and that and faster CPU.

My 2 years old Mac mini m2 pro is fine, no plans to buy a new one, no need to, if I uised it for business and loads of video editing, then maybe, but I don't.
People need to learn to just say, nope, don't need it.
 
Apple needs to stop improving their silicon every year because it's angering too many people.
Why on Earth would offering more performance each year at the same or lower price anger people? Don’t understand that at all. Nobody makes you upgrade until you need to.
 
  • Like
Reactions: wyrdness
Genuine question... why? I have an M1 and would like to see how it compares with the M5, why would you start the comparison at the M1 Max?

Well, that’s a good question — because it’s the first M-series MacBook Pro.

Also, the M1 Max is an exceptionally good computer for its age. The M1 is not outdated at all yet, so it still makes sense to compare the current lineup to it. I think the M5 is actually the first one that will have a real edge compared to the M1 (16" version).
 
So Wyrdness, you didn't like my post. I saw earlier that you stated the single-core performance was 15% faster. Where did you pull that statistic from the article? My comment pointed to the observation that the article didn't mention the single-core stats. But I'm old. I could have missed it. Please quote the line in the article that mentions 15% improvement for single-core, and I will be humbled (won't be the first time).
 
Compared to the M4 chip that Apple launched in May 2024, the M5 delivers:
  • 27.5% higher unified memory bandwidth

In addition to general performance claims, Apple published a set of specific real-world workload results showing measurable gains in AI-driven applications:

  • 4×+ peak GPU compute performance for AI
  • 3.6× faster time to first token (LLM)

M4 Chip M5 Chip
120 GB/s unified memory bandwidth 153GB/s unified memory bandwidth

For users whose workloads include on-device AI inference, complex 3D rendering, or other GPU-bound or memory-intensive tasks, the jump from M4 to M5 is material. The combination of per-core Neural Accelerators, higher memory bandwidth, and new GPU architecture produces multi-fold speed-ups in certain AI operations. In environments where time-to-result directly affects workflow, such as local LLMs, diffusion models, video enhancement, or ray-traced production or gaming, the M5 represents a meaningful step-change rather than a minor iteration.
I guess if you are sticking with a laptop (M1-M4) and did not have 32GB, this new one would get you a bit more local LLM performance. You would really have to know the exact models you use in your workflow to know if the extra RAM and bandwidth would make a big enough difference to upgrade. But if you needed it, you would probably choose to get a M4 Max MBP with with faster memory bandwidth and potentially more RAM.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.