Performance Comparison: M3 Max MacBook Pro vs. M1 Max MacBook Pro

[MacRumors]

Apple last week launched its new MacBook Pro models with M3, M3 Pro, and M3 Max chips. We already took a look at the M3 Pro MacBook Pro, but we also wanted to test out the top-of-the-line M3 Max chip to see how it measures up.

Subscribe to the MacRumors YouTube channel for more videos.​

While the M3 Pro's performance was disappointing overall compared to the M2 Pro and even the M1 Pro, the same can't be said of the M3 Max. Using Geekbench, the M3 Max is about as fast as the M2 Ultra, earning a single-core score of 3217 and a multi-core score of 21597. The M2 Max has a single-core score of 2737 and a multi-core score of 14503, and the difference is even more pronounced compared to the 2021 M1 Max, the first 16-inch MacBook Pro to get an Apple silicon chip. The M1 Max has a single-core score of 2379 and a multi-core score of 12206.

Other benchmarking tests show similar major improvements between the 2023 M3 Max and the 2021 M1 Max. In Cinebench, the M3 Max earned a multi-core CPU score of 1601, compared to a 788 score from the M1 Max. A classroom render in Blender took three and a half minutes with the M3 Max machine, and eight and a half minutes with the M1 Max machine.

Exporting a 16 minute 4K video with multiple effects took five and a half minutes on the M3 Max, and seven and a half minutes on the M1 Max.

As for SSD speeds, those were about the same. The M1 Max saw read/write scores of 5727/5980, respectively, while the M3 Max had read/write scores of 5032/6197, respectively.

You'll want to watch our full video for a more detailed comparison between the two machines, which gives a useful look into just how far Apple silicon has come in two years. The M1 Max is still a super fast chip and you're probably not going to want to upgrade to the M3 Max if you've already got an M1 Max machine, but in some tests, the new chip is up to twice as fast.

Article Link: Performance Comparison: M3 Max MacBook Pro vs. M1 Max MacBook Pro

 

[GooseInTheCaboose]

Is there any reason to believe the ssd speeds on the m3pro are slower?

And is there a detailed m3pro test article yet?

 

[mdhaus72]

Pretty amazing results overall.

 

[mecloud]

Is there any reason to believe the ssd speeds on the m3pro are slower?

And is there a detailed m3pro test article yet?

I tested the base 512gb M3 Pro, and yes the internal SSD speeds were somewhat slower, around 4,000mb/sec. But that was still an improvement over the 512gb M2 Pro, which measured around 2,800mb/sec. On my M3 Max (base config of the high end CPU/GPU), I’m seeing about 6,400mb/sec write, 5,100mb/sec read.

 

[Born Again]

Small gains… Apple just did such a great job with the m1max you don’t need to upgrade.

Even MKBHD cancelled his m3max macbook order. To me that speaks volumes.

 

[Timpetus]

Good to see the improvements, though I don't need the extra power over my M1 Max. Can't wait to see what they do have on offer by the time I upgrade in about 5-6 years!

 

[dcingie]

Let’s compare it against the original iPhone next guys

 

[Red Oak]

Is this now when I complain on MR that M3 is not fast enough vs. M1, as I sit on my couch eating Cheetos and watch Netflix?

 

[mapsdotapp]

Any word on how the base M3 compares to base M1 and M2 chips?

 

[Saturnine]

I have an M1 Max MacBook Pro from launch in 2021, so it’s just about two years old.

I went for 64GB RAM and a 2TB SSD, which still seems to be a decent spec.

I see that the M3 Max appears to outperform my machine by about 30%. Even so, I’m still more than happy with its performance and I expect I’ll get another two or three years out of it before it’s handed down to a family member.

 

[daneoni]

The numbers look impressive but 2 mins real-world gains...factoring the cost does not seem like true value.

Again if you're coming from Intel then by all means. But if you already have an M series chip then really there's less incentive

Unless you have an esoteric workflow thats takes advantage of every minutae of performance gain

...or just have have disposable income.

 

[GuruZac]

Interesting. The M3 Max is really the big jump up. That color looks spectacular in videos.

 

[mecloud]

Any word on how the base M3 compares to base M1 and M2 chips?

Base M1 Pro 512gb drive speed is faster than both the base 512gb M2 Pro and M3 Pro (all 16 inch) from my testing. Yes, weird. Measurement of my base M1 Pro is 4377.9 write, 5481.5 read according to my saved Blackmagic screenshot. The M1 Max I had (4TB SSD) was faster, but I don’t recall exactly by how much.

 

[mecloud]

The numbers look impressive but 2 mins real-world gains...factoring the cost does not seem like true value.

Again if you're coming from Intel then by all means. But if you already have an M series chip then really there's less incentive

Unless you have an esoteric workflow thats takes advantage of every minutae of performance gain

...or just have have disposable income.

In my case, it all started 4 weeks ago today, when my M1 Max did the “black screen on startup” thing, which turned out to be a Mac OS bug in Monterey 12.7 and Ventura 13.6. It took me quite a bit of time to realize I was not dealing with hardware failure as it originally appeared, but the Mac OS bug that was addressed in Ventura 13.6.2 last week.

If none of that had happened, I’d still be using the M1 Max and wouldn’t be considering upgrading.

In my use, there’s benefit to the M3 Max versus the Pro. That’s now my primary laptop, with the base M1 Pro my backup.

 

[Klagarde]

While we would like to see bigger improvements since the node sizes became smaller, compared to Intel in the past 10 years the changes per generation are still big.

 

[Warped9]

Max Tech compared the M3 Pro MacBook Pro to the M2 Pro MacBook Pro and the results were surprising. Geekbench tests don’t tell the whole story. Suffice to say the M3 Pro MacBook Pro did quite well and under certain use cases it is indeed worth considering the upgrade.

 

[d4cloo]

Even pro users don’t often hit peak performance.
It is more valuable to have a snappy machine during productivity (e.g a fast 3D viewport with decent rendering in Blender, a snappy FCPX timeline) compared to a final render.
Final renders don’t happy often. Usually you rely on a proxy quality when authoring content and only at the end you do an actual render.
Only when you play games, you want sustained, trustworthy (GPU) performance. However: Mac 🤷🏻‍♂️

Which makes the new M3 less attractive for existing Apple Silicon users.

 

[mecloud]

Which makes the new M3 less attractive for existing Apple Silicon users.

From watching the introduction a couple weeks ago, they were really hammering on the whole theme of “Intel Mac users, time to upgrade!”. I got a very strong hint that Intel support goes away in the next major Mac OS version.

I agree that unless someone is really pushing the limits of existing Apple silicon hardware (maybe looking to upgrade product tiers because of that), the M3 MBP’s are a pretty incremental upgrade from the end-user standpoint.

 

[lk400]

Is this now when I complain on MR that M3 is not fast enough vs. M1, as I sit on my couch eating Cheetos and watch Netflix?

You could always go back to your day job of making professional looking videos about the tools you use to make professional looking videos.

 

[Gloor]

I think Dan needs to learn what Blender is used for as his explaining is not that great.
Saving 5 mins on a single image is HUGE saving. If you render animation and each second has 24 frames (film) you are saving 2!!! hours per second of footage.
Its unlikely that you do this on a production thing, but for students or small freelancers, that is a BIG deal.

Talking about 5 mins a day (as Dan put it) is really not the right comparison as its never just a single frame.

 

[ChrisA]

Is this now when I complain on MR that M3 is not fast enough vs. M1, as I sit on my couch eating Cheetos and watch Netflix?

Yes, it took me an hour to watch that video on my M1. I'd expect to see the same video in 30 minutes on an M3. But no, it is still taking an hour. Same with emails, I'm not reading them any faster with the M3. What's wrong Apple?

Seriously, why don't these reviewers focus on real-world tasks that real people do? Who cares how fast a benchmark runs.

 

[3xBoom]

50 shades of Space Gray

 

[ChrisA]

I think Dan needs to learn what Blender is used for as his explaining is not that great.
Saving 5 mins on a single image is HUGE saving. If you render animation and each second has 24 frames (film) you are saving 2!!! hours per second of footage.
Its unlikely that you do this on a production thing, but for students or small freelancers, that is a BIG deal.

Talking about 5 mins a day (as Dan put it) is really not the right comparison as its never just a single frame.

Yes, this is what has to happen in these reviews. The trouble is the bloggers don't know much about how computers are used so all they are really qualified to do is download a benchmark app and click on it. A real journalist would go out and find professionals who do use computers for work and interview them about what they do and then have different kinds of them try the new computer and summarize how each kind of use case could benefit or not from an upgrade. But this kind of original research takes hours and days and weeks while clicking a benchmark app takes only a minute or two.

That said, blame this on the readers (you and I) we don't want to pay for quality content and prefer free crap.

 

[Warped9]

 

[3xBoom]

Let’s compare it against the original iPhone next guys

Well, you are closer to the truth that you intended with your snipe comment - since Apple clearly pitched m3 to the intel crowd, the proper comparison should have been with the last Intel MBPs and not m1 or m2. Apple has a crystal clear knowledge of the size of the installed base still using MBPs running of Intel chips and overheating the world and the goal of m3 is move those hundreds of thousands of users to Apple silicone. It was clear to those with half a brain in 2020 that Apple silicone will need several years to mature both as chip and as an environment. Likely. few of those "rocket scientist" decided to commit to the change at that time but now, three years down the road, the time is right.