Apple unveils M3, M3 Pro, and M3 Max | Hacker News Reader

Apple unveils M3, M3 Pro, and M3 Max

apple.com

1,035 pointsehPReth1,126 comments

Related ongoing threads:

Apple unveils the new MacBook Pro featuring the M3 family of chips - https://news.ycombinator.com/item?id=38078065

Apple supercharges 24‑inch iMac with new M3 chip - https://news.ycombinator.com/item?id=38078068

A few things I noticed, as I'm seeing the variety of SKUs becoming more complex.

- Note that memory bandwidth is down. M2 Pro had 200GB/s, M3 Pro only has 150GB/s. M3 Max only has 400GB/s on the higher binned part.

- Just like the low-spec M3 14" has one fewer Thunderbolt port, it also doesn't officially support Thunderbolt 4 (like M1/M2 before it)

- The M3 Pro loses the option for an 8TB SSD. Likely because it was a low volume part for that spec.

- The M3 Pro actually has more E-cores than the Max (6 vs 4). Interesting to see them take this away on a higher-specced part; seems like Intel wouldn't do this

I'm wondering now if Apple tracks this sort of stuff in their released machines. I know my iPad always asks me if I want to send stats to Apple (and I always say no). So let's say that enough people do, then do they have a good idea of how often all the performance cores are used? Max memory B/W consumed? Stuff like that. Back when I was at Intel there were always interesting tradeoffs between available silicon/thermal/margin resource and what made it into the chip. Of course Intel didn't have any way (at that time) to collect statistics so it was always "... but I think we should ..." not a lot of data.

In addition to the reduced memory bandwidth, the M3 pro also loses 2 performance cores for only 2 more efficiency cores.

M2 pro: 8 performance cores + 4 efficiency cores.

M3 pro: 6 performance cores + 6 efficiency cores.

Not a great trade... I'm not sure the M3 pro can be considered an upgrade

> Just like the low-spec M3 14" has one fewer Thunderbolt port, it also doesn't officially support Thunderbolt 4 (like M1/M2 before it)

I believe this is due to the TB4 spec requiring support for two external displays on a single port. The base spec M series SoCs only support one external display.

I’d expect the ports to work identically to a TB4 port in all other aspects.

> seems like Intel wouldn't do this

Wouldn’t do what? Intel has more E-cores than P-cores on most of their range, and especially on the higher end e.g. on raptor lake S the i9 all have 16 E and 8 P, the i7s have 8:8, only the lower end of the i5 (below 13500) have more P than E cores. And the i3 have no E cores.

The story is somewhat similar on mobile (H and HX), a minority of SKUs have more P than E, and none of them in P and U.

In fact that was one of the things which surprised me when Intel started releasing asymmetric SMT, they seemed to bank heavily on E cores when mobile and Apple had mostly been 1:1 or biased towards P cores.

The SKUs are becoming more complex because they are probably learning why Intel/AMD have so many SKUs. Making complex chips at scale results in a range of less-than-ideal chips. This drives a the need to segment and bin chips into different SKUs to reduce losses, rather than trying to sell one SKU and throw awaying the anomalies.

> Note that memory bandwidth is down. M2 Pro had 200GB/s, M3 Pro only has 150GB/s. M3 Max only has 400GB/s on the higher binned part.

Is this because they are not populating all the memory channels, or just using lesser memory ICs?

If its the former... thats annoying. It just makes their products worse for artificial segmentation and very little cost savings.

The new M3 14" MBP seems to be a red herring - why does it even exist? Why not just refresh the MBA instead?

An obvious rule of thumb is for "Pro"-branded laptops to only use "Pro"-branded chips. It's what they follow for the iPhone lineup, but I suppose iPad Pros also use non-Pro chips. Just seems like a very confusing SKU to create, and definitely something Steve wouldn't approve of.

It was pretty hard to saturate the memory bandwidth on the M2 on the CPU side (not sure about the GPU).

> The M3 Pro loses the option for an 8TB SSD. Likely because it was a low volume part for that spec.

That's not super surprising to me. Apple loves to charge stupid prices for storage and memory. Maybe it's worth it for lots of people to have the convenience of built in storage at the lower levels, but I have to imagine that most people would want 8TB of SSD would rather just get an external solution for... much less.

I think what Apple is pushing for is computing efficiency. It still gets faster but with much less power. Focusing on performance solely would be the wrong way to evaluate these chips. https://www.tomsguide.com/news/apple-m3-chip

> Note that memory bandwidth is down. M2 Pro had 200GB/s, M3 Pro only has 150GB/s. M3 Max only has 400GB/s on the higher binned part.

just contrasting this with the recent TR4 announcements from AMD, apparently their PRO variants top (theoretically at least) at around 325GB/s (non-pro versions are half of this), so just from that perspective alone M3 Max's might be better ?

i always have the naive assumption here that keeping the-beast i.e. the cpu fed with data is much better for overall performance than just clock-rates etc.

The 2x USB/Thunderbolt ports are on the same side. :(

Unfortunately, I don't see apple doing anything but price discrimination using mostly number goes up.

The max is probably only going to be in desktops, so better to use the die area for other things than E cores

Wish that you could get the 16-core CPU with the smaller Max GPU, but alas I just ordered one anyway.

One thing that left me a bit confused was the comparison against Intel Macs. Although I am still using an Intel 16-inch MacBook, I really wanted to see how the M3 fared against the M2, not Intel and M1. I think it's no surprise the M3 exceeds the Intel Core i7-9750H in basically all of Apple's own benchmarking. My real question, which will probably be answered next week, is how it compares to the generation right before it.

My work laptop is a 14-inch MacBook Pro and I've been impressed with the battery life considering all of the containers that run (on Kubernetes!) as part of my dev workflow. I guess Apple deciding to compare Intel and M1 was to try to convince existing MacBook users to upgrade to the latest generation CPUs.

I imagine the amount of people upgrading from an M2 Mac will be close to 0, while there's a lot of Intel stragglers out there.

For years now, Apple's biggest competitor is Apple from five years ago. One of their biggest threats is that sales flatline because people are still using their perfectly good laptops and phones from several years ago. And that's a hard problem to deal with when you advertise your products as high-end goods that will last a long time.

The comparison with M2 is not very useful at this point. Throughout the presentation, the gains were something like 15-20% in some areas.

Even if they do their own chips, Apple cannot achieve a revolutionary performance gain year over year, that's why between two generations we'll see something like this for a long time. And since their idea is for people to be using Macs for many years, it makes sense to compare it to older generations to try and persuade people to do the upgrade.

Not just that, but if you're not running ARM images then it's running a VM for your docker images to boot. It's just mind-blowing how many containers I have running and my laptop is cool to the touch, and at full battery.

I was more surprised not to see any AI-specific benchmarks—sure, the most popular open-source models are from avowed competitors, but there should have been a way to define a re-training task that would be relevant for the wave of ML programmers.

There were comparisons to the M1 and M2 in the presentation, and indeed the linked website.

I still know a number of my colleagues who haven't jumped onto the Apple Silicon machines despite our 2019 Intel machines being out of warranty (and thus eligible for replacement) for two years now

I think this is squarely aimed at people who are holding onto their Intel-based Macs with an iron grip. There are always people out there that don't ever want to move to another architecture. I saw it going from Motorola 68030s to PowerPC. I saw people not wanting to upgrade from PowerPC to Intel. Now we're still seeing the people who don't want to migrate to Apple Silicon. They may have legit reasons and what-not. But time is ticking.

So I think it's mostly aimed at the Intel hold-overs.

In case you are not already using orbstack, that might give you even more battery life (not affiliated, just a fan).

A container that is idle and do not serve any request only really as memory in its footprint. There is no reason having tens or hundreds of processes not occupying cpu and i/o time would affect battery life in a significant manner.

It shows the comparison right on the chart, what else you need? But you should get m3 most times if you want to use it for the longest period.

Out of curiosity, how much RAM do you have on your MacBook to run containers?

I feel very disappointed that the laptops lack decent i/o. They spend all this time talking about how advanced the GPU is, but it's a $1,600 laptop that can only power one external display or a 2,000 laptop that can only power two displays?

Plus they are nerfing the desktops by only offering the all-in-ones with m2, so you can walk into a store and want to buy any other desktop now you are paying a huge premium for outdated tech :(

I was really hoping for FaceID and better display support, but I don't feel there was any compelling reason to upgrade anything. If you preferred Windows before, you probably still prefer Windows. If you have an M processor there is no compelling reason to upgrade. If you were thinking about getting a Studio or a mini, you should probably just wait until they update the processor.

> I feel very disappointed that the laptops lack decent i/o. They spend all this time talking about how advanced the GPU is, but it's a $1,600 laptop that can only power one external display or a 2,000 laptop that can only power two displays?

yep. powering 2+ screens on a mac is not a straightforward process. for the prices Apple is charging on their "pro" machines, it should be simple to plug in 2 or even more screens to get work done.

> Plus they are nerfing the desktops by only offering the all-in-ones with m2

I'm confused by this statement. The iMacs leapfrogged M2. The new models shipping next month will have regular M3 chips, replacing M1 models.

The single external display driver can drive more than one display at resolutions at 4k or lower. You can daisychain two 4k monitors on a Mac that has only one external display driver when you use the Displaylink adapter (it’s a driver). It somehow internally reuses the same display output for both displays.

The USB-C can power one 6k or 2 4k monitors.

The HDMI port on M2 Pro and newer can power one 8k.

An interesting thing about the M2 Pro and M3 Pro is that they shifted away from being mostly performance cores. The M1 Pro was either 6+2 P+E or 8+2 P+E. The M2 Pro was either 6+4 or 8+4. The new M3 Pro is either 5+6 or 6+6.

Apple has been shifting away from performance cores with each generation. M1 Pro: 75-80% P-cores; M2 Pro: 60-67% P-cores; M3 Pro: 45-50% P-cores.

This shows up when you look at Geekbench results. A 10-core M2 Pro (6+4) gets 12,100 while a 10-core M1 Pro (8+2) gets 12,202. The 12-core M2 Pro (8+4) gets 14,221. That's a 16.5% increase from having 20% more cores. In some ways, this feels like an odd result. Adding two additional M2 P-cores gets Apple a comparatively small result over the 10-core M2 Pro (less than the average core performance). However, adding two efficiency cores over the M1 Pro gives them the same 16.5% boost over the 8+2 M1 Pro.

If I had to guess, maybe it's thermal throttling when running the benchmark. If the additional P-cores can't truly be P-cores under 100% load, then their impact shows up less. Likewise, if the E-cores can match P-core performance under heavy thermal load, then it could show up as being roughly equivalent in the benchmark.

I wonder if real-world scenarios end up differing from benchmarks in a meaningful way around this. For example, core-pinning can be useful for warm caches and in a real-world scenario you might have a process the OS tries to pin to P-core-1 that has spikes in utilization while another process is pinned to P-core-2 with similar spikes. So both get the performance of a P-core and warm caches while the thermal load isn't that high so the P-cores are still at their peak performance (unlike a benchmark that's trying to use all cores as much as possible at the same time).

Maybe this is a business decision more than something based around how the chip performs. The big selling point of the M1/M2 Max was graphics (maybe the extra RAM as well). You could get the same CPU in the Pro or Max. Now the M3 Pro is a 5+6 or 6+6 CPU while the M3 Max is 10+4. 67-100% more performance cores becomes a selling point for the M3 Max even for those who might not care about graphics as much.

I think your last paragraph sounds about right. The Max is a compromise on battery life for those who want the absolute most possible performance, but I think Apple is seeing a much wider audience get the Pro chip. Anecdotally, whenever I see a tech person mention their Apple silicon device, it's almost always a Pro chip. I'm sure only a small fraction of that user base is hitting all cores with a full load on a regular basis, but a majority are certainly enjoying a killer battery life.

It seems like Apple is learning from Intel here.

Apple made their E-cores a lot faster the past couple generations. Despite being a tiny fraction of the size, they get nearly 50% of the performance (yay diminishing returns) while using around 10x less power.

If your task only scales to 1-4 cores, you need those cores to be fast, but generally speaking, if you can scale more than that, you can probably scale to a LOT of cores. From this perspective, 6-8 cores allow a couple of those lightly-threaded applications at the same time which is usually all the more you'll be running.

After that, putting 4-6 E-cores+cache in the space of just one P-core+cache means you're still using half the power, but getting a lot more performance in those very scalable applications.

AMD is headed this same direction with Zen 4c and 5c chiplets where you'll have one performance chiplet and multiple efficiency chiplets.

The area and power wins are just too big to ignore.

> maybe it's thermal throttling

The GPU is a very large fraction of the thermal output these SoCs are engineered to handle, so there really shouldn't be any throttling if you're only loading the CPU.

Significantly easier to get benefits from scaling GPU cores than CPU cores.

What's with the 36 GB option? The other memory configs (16 GB, 64 GB) are still clean powers of two. Size suggests that they're using ECC-capable memory but using the extra width intended to support ECC for data... but why would this only apply to a single size? Part availability?

EDIT: Digging this a bit, it's not (one or more) 72-bit wide busses with 2^32 words as I'd expect as a gray-beard, it's (probably) six 32-bit wide busses with 6 GiB per bus; and this use of 1.5 * 2^N deep memories has become relatively common with the use of IC stacking, with 12 power-of-two sized ICs stacked in a single package (instead of the more "comfortable" 8 high or 16 high stacks of the same ICs giving powers of two).

They've been doing more multiples of 6/12 lately, I supposed it's related to the available chip sizes for the LPDDR5X.

M2 and M3 go up to 24GB

M3 Pro is available in 18GB and 36GB

M3 Max is available in 36GB, 48GB, 64GB, 96GB, and 128GB

It also appears they've dropped the memory bandwidth from 200GB/s on the M1/M2 Pro to 150MB/s on the M3 Pro, and you have to upgrade to the tippy top M3 Max chip to get the full 400MB/s bandwidth experienced on the M1/M2 Max chips.

It has nothing to do with ECC. They are using LPDDR5X memory, and if you follow Android phones these number shouldn't be surprising.

I prefer that tbh. If I'm frequently saturating 16GB of memory in the previous gen and it is time for an upgrade, I probably don't need literally double the memory, but 24GB would be nice.

There’s also a 18GB M3 Pro config of the MacBook Pro. Very unusual.

64 GB chips that failed testing? That's one way to increase yield.

Clive Sinclair did that trick back in the 80s.

It looks like these models won't be as useful for LLM inference which are heavily memory bandwidth constrained. The Macbook Pro page shows M3 at 100GB/s,150GB/s, and 300GB/s vs M2 at 200GB/s and 400GB/s. 400GB/s is available for M3 if you opt for the high gpu config, but interesting to see it go down across all of these models.

The presentation mentioned dynamic GPU caching: that seems like something transformer models would like.

I don't understand why they don't update the entire lineup whenever a new chip comes out, especially if really only swapping the chip in the macbook pro and imac here.

On a separate note does anyone have any insight on how unified memory compares to vram in the context of machine learning performance? Considering an H100 with 80GB costs like $30k a maxed out macbook pro with 128gb unified memory for $5k is interesting. Is it remotely comparable or compelling for large models considering realistically most prosumers are capped at the 24gb nvidia cards for anything resembling a reasonable budget.

I expect it's a combination:

1. Engineering bandwidth (it's work to upgrade a lineup! They probably need at least a year)

2. Manufacturing bandwidth - they probably spent the first half of the year manufacturing the 3nm iphone processor while researching the mbp processor. I expect it's difficult to ramp manufacturing on many chips simultaneously

3. Sure, consider the demand side. If my parents are going to buy a new laptop and a new phone this year, I think they would be more likely to do that 6 months apart. Similarly, I expect keeping a cadence of announce products a,b,c in this quarter and d,e,f in that quarter helps keep apple in the news (for good/exciting things)

Wrt machine learning: I can't wait for the results to come out for this once we can get our hands on it

Speculating, but I'd wager it's to clear out existing inventory by phasing out the old chips on other SKUs over time. E.g., you start with the expensive Macbook Pro models when you have fewer M3 chips in production, since those will sell slowly, and then you clear out your remaining M2 inventory in the Air lines while waiting for a "refresh" when M3 production is more ramped up.

I have no idea what Apple's sales are for the macbooks or if Airs sell more than Pros, etc., but they're definitely making profit this way to be this consistent in their approach.

Besides wanting to clear out existing M1/M2 capacity, I don't there's enough M3 stock right now.

I forget how many fabs TSMC has that have 3-nanometer capability, but it's gotta be low. Apple's supposedly got _all_ of TSMC's capacity[0], but I doubt it's enough.

[0]: Apple is saving “billions” on chips thanks to unique deal with TSMC | https://news.ycombinator.com/item?id=37040722

They just updated their macbook air and mac studios and mac pros like 6 months ago? If they updated right now, the previous owners would feel utterly screwed over.

The M2 Ultra tops out at 800GB/s unified memory bandwidth. A 4090, on the other hand, has 1,008 GB/s. On the PC side, dual-channel DDR4-6400 offers a bandwidth of 102 GB/s.

Depends on if you're talking about training performance or inference performance. It's probably a decent deal for inference on such large models, but I doubt it's anywhere near competitive for training. In the 'consumer' space there's also the A6000 with 48GB for ~$4k.

It's probably due to a combination of the processor manufacture rate and the quarterly sales boost they get throughout the year

It makes a world of sense if you think of it from a logistics point of view, especially considering Apple sells insane volumes of everything they sell.

It costs less to manufacture the current generation ship and the last generation chip concurrently, it allows them to push the newest generation chip out the door faster without running out of supply or having to open up excessive production lines.

>H100 vs apple silicon

while it's only one spec, H100 has over 2TB/s of memory bandwidth and Apple silicon caps out at 400GB/s. H100 in general kicks the pants out of Apple silicon performance and has far more support. These parts aren't really in the same league but Apple Silicon has become a popular hobbyist choice for LLM inferencing.

Also see: The Jetson AGX Orin 64gb, Nvidia RTX A6000 48gb, AMD Radeon Pro W7900 48gb, two nvlinked 24gb 3090's.

Regarding updating the lineup, I would guess that it is completely intentional to keep the carrot just out of reach across the lineup. Each new chip release rotates the product categories in just the right way to keep the consumers consuming.

Right? I’ve just setup my office and for now I’m using my MBP hooked up to a couple of Studio Displays - which I adore for software development. I was planning on picking up a Mac Mini or Mac Studio so that I could still have the MBP spare.

Maybe not. It’s a bit of a pain having to unplug and replug everything in, not to mention having windows and such shuffle around each time.

My gut tells me the Macbook Air sales have been cannibalizing Macbook pro sales.

For the M1 and M2 generations, the airs released in the fall and the pros a few months later in the spring. The airs have gotten so good that many people that would have gotten the more expensive (and more lucrative) pros got airs instead.

> Additionally, support for up to 128GB of memory unlocks workflows previously not possible on a laptop, such as AI developers working with even larger transformer models with billions of parameters.

The presentation explicitly highlighted this very randomly, without showing how AI development actually works.

I know Apple has put more effort into Apple Silicon support for PyTorch but is it there yet?

I don't think it is? But I would love to be wrong. Cheap mobile Nvidia cards outperform regular M2 chips on small learning tasks.

M3 Max might be a contender, but Metal is far from CUDA yet? A laptop M3 Max looks like it can compete with a Nvidia 3070.

It's an unfair comparison as Nvidia is a huge desktop heat/energy hog and Apple M chips are really efficient. Let's see how desktop M3 chips fare.

I'm pretty sure pytorch supports it OOTB. In my experience it works even better than ROCM

Are the cool kids using GGML?

It's quite impressive that the event was filmed on iPhone 15 Pro[1], although it involved professional lighting and various equipment, which is not typical for the average user.

With the ability to capture in Log, it's possible that the next iPhone release might be filmed using the very phone that's being unveiled.

[1] Source: https://www.youtube.com/live/ctkW3V0Mh-k?t=30m02s

Really love the pace of Apple innovations. Nowadays, I am biased towards battery and weight (MacBook Air) if it is for portability and the other line of notebooks are converted into desktops via TB exactly because of weight and battery life. I feel the difference in battery use from one to the other. The Air seems like it has "infinite battery" evn when the others use Mx processors.

Also waiting for a great Linux port battery/weight wise.

i’ve had the 16gb ram macbook air for 3 years now and it’s still an incredible machine considering lightness, battery, and passive cooling.

I was seriously considering an m3 pro for the space black color until I confirmed it wasn't passively cooled. An extra monitor would be nice at work but the air is just too great as a laptop.

Interesting that the Myst remake is being used specifically as an example, with how the original was created on Macs using HyperCard.

Yeah, I thought that was a little poetic :)

Especially since (and this just occurred to me), in a roundabout way the original may have been the first Mac game to ever use raytracing (in its pre-rendering)

> Myst

I was just coming here to ask about the Myst reference, since I played that 25 (?) years ago and wondered why they mentioned it. So there's a remake?

That one, and Sketchup really confused me.

I think the M series laptops were the first full manifestation of the laptop idea.

Instant sleep/resume, great input devices and monitor, cpu power at will without overheating or loud fan, super fast shock resistant storage, 20+ hours of battery life, no battery drain when off.

It took us 30 years but we finally have it.

Now let’s hope WIntel can also catch up soon.

It's kind of surprising that Intel (or I guess x86 in general) still has basic issues with sleep, laptops waking up in your bag and that kind of thing. The forced modern standby thing they're doing has only made it worse.

It’s kind of strange their benchmark for “Machine Learning Programmers” is “Simulation of Dynamical Systems in MATLAB.” Seems like they could have capitalized better by using a generative ai inference benchmark.

iPhone 15 Pro NPU can do 35 TOPS. M3 can only do 18 TOPS. That's why they didn't focus on ML in the video.

Maybe they know M3 is not up for that kind of task.

These are still going to be used in laptops with soldered flash chips for the SSD. Apple's laptops aren't meant to be used by people who expect to repair their laptop when the storage fails. Many people don't have access to replacement chips and services to have the flash chips found on the board replaced. It's as if their hardware is meant to be e-waste.

The RAM is still not ECC. They focus so much on the shallow marketing without shipping anything which makes a difference.

My Windows desktop has generated significantly more e-waste per year than the Macs I have owned. Full stop. I've upgraded SSDs from 500gb to 1TB to now 2TB, upgraded graphics cards, and replaced a massive broken aluminum and copper heatsink. All in the last two years.

I've had the same MacBook, in a similar timeframe. It gets just as much use. Its still rock-solid.

Your concern is a hypothetical one. The inability to upgrade has, in a VERY REAL sense, meant for me: I overbuy specs upfront, so I don't have to upgrade (and thus, generate e-waste). It also experientially means that the machines are more reliable. The least wasteful machine is one that doesn't have to be upgraded; not one that can be and needs to be. The 500gb of storage I have in my MacBook sometimes feels limiting; but the cost of upgrading (a whole new machine) has stopped me from actually going out and doing it; and thus less e-waste is generated.

Eventually I will brainstorm what to do with this machine once its outlived its useful life as a laptop. I'm thinking, server rack. We've got Asahi, its got thunderbolt so wiring up super fast storage drives is a cinch.

> It's as if their hardware is meant to be e-waste.

But the chassis is made from 100% recycled aluminum!~

On a serious note, Apple has a simple and working strategy, to which they're 100% committed: repair only in authorized service points, otherwise recycle. Making devices user-repairable adds costs and compromises on specs; you can't put a number on the disks being replaceable as easily as you can put a number on size, IO bandwidth, or number of write cycles, and numbers is what looks good on benchmarks.

You also have to consider Apple's scale: they have literally billions of deployed devices (everything from AirPods to Mac Pros) that they need to support, so what could work for a different vendor won't necessarily work for them.

(I don't necessarily agree with that strategy from end-user POV, merely pointing out the context.)

The point is, they don't want people to fix or upgrade their hardware, and they will fight to defend this strategy. Give me one reason why they should to it, from their perspective. If they have them glued, the equipment has shorter life and customers need to purchase a new one quicker. And since they can't replace components themselves, they have to pay exorbitant prices for 16 GB RAM or 1 TB storage that is cheaper than ever.

These are nice machines, though, and I buy one for building iOS apps every couple of years - but usually 1-2 generations later so that the pricing is more reasonable.

Average error rate in non-ECC consumer DRAM is in the order of one in 100 million, which is not insignificant.

Strange that they mostly compare performance to M1, not M2. Probably means it's not as much faster than M2 as they would like.

Its on the release linked. 15% faster than m2 for P-cores and 30% for e-cores, which is misleading, because it has 2 more of them so they're not necessarily that much faster.

For me, still on m1 is good news, one more generation I can skip without feeling too much FOMO. Graphics improvement are irrelevant since im not gaming on them and cpu difference is fine for being just <40% overall.

Exactly. I imagine most of the benefit of the new 3nm process is related to power draw. I'll bet the memory is the same stuff as the prior generation though, given those specs haven't seemed to change at all.

Don't forget that there are no M2 iMacs

They compared it to both, and most people upgrading will have an M1 so it's the more apt comparison (and of course, a bigger number).

What's the deal with "dynamic caching"? I wish they would have talked about it more. It sounds like they're reducing memory allocations by only allocating what is actually needed.

You got it. Instead of letting software developers allocate a static cache with overhead that eats a little into the unified memory, the Mac gets to decide how to dynamically allocate and release cache for graphics. They said it would be "transparent to developers," but not really sure what that means exactly.

From what I understand this is about assigning GPU resources (such as register files and other on-core memory) to shaders. Imagine you have a complex shader that has two paths, a frequently taken fast one that needs X bytes of on-chip memory to function and a rarely taken slow path that needs Y bytes (Y > X). Usually this stuff is statically partitioned, so you have to provision Y bytes for the slow path to run this shader. With dynamic partitioning, the shader will only allocate Y bytes if the slow path is hit, which frees the resources to load up more concurrent shader programs and improve the shader occupancy.

This stuff is only really relevant if you are dealing with complex uber-shaders and recursive function invocations, both of which are fairly common in raytracing pipelines.

They kept mentioning performance relative to intel Macs which makes me think there is a large cohort of people sticking to their x86 rigs due to compatibility. Being able to run an x86 linux or windows VM is still a requirement for me.

It's only been 3 years since the M1, most people aren't buying new computers that frequently.

Windows for ARM runs well on Apple Silicon, and has its own translation layer for x86 software. It should be fine unless you need specific x86-only drivers.

I'm still using mine. Parallels can only MDM enroll the mac VM on the intel device. Apple silicon doesn't support the feature of changing the serial number. So for testing mdm profiles I have to use it or carry two devices around.

I'm also still running an i9. I can't justify buying a new one since my 2019 model is still running strong!

Who knows but my guess is that the x86 requirement is probably for a minority of users.

Lots of people are probably sticking to Intel Macs simply because the averege user doesn't care about performance and will keep using a computer until it dies.

https://mac.getutm.app runs just fine with x86 linux vms. I'm running a few ubuntu ones as we speak. It's painless and quite reasonably speedy.

And still starting at 8GB of RAM… Disappointed.

and yet at this point pretty much expected. So obnoxious and odious - the reality is for any C/C++/rust or similar compilation you need a minimum of 1gb per compilation core for compilation alone, before you consider the rest of the OS :-/

You're saying people shouldn't be able to buy 8GB RAM Macs, which is crazy. I had to check how much memory my laptop has, and it's 8GB and I've never noticed.

I guess you're just complaining about what you get for the lowest price.

Seriously.... My 16gb MacBook pro M1 work machine runs out of memory constantly. I had a MacBook air 8gb and it was useless, it would freeze constantly and have crackly audio. I found that 24GB ram is really the minimum that is usable on M-series when doing any type of basic work.

Likely an output of so many electron based apps which are memory hungry, but if the memory is supposedly "really fast" then I would expect better performance. Never had the issue with Intel.

But with a nice bit or memory, the performance is really nice.

> With the M3 family of chips, hardware-accelerated ray tracing comes to the Mac for the first time. Ray tracing models the properties of light as it interacts with a scene, allowing apps to create extremely realistic and physically accurate images. This, along with the new graphics architecture, allows pro apps to deliver up to 2.5x the speed of the M1 family of chips.

Does this mean we can finally play Cyberpunk 2077 on a laptop? Is it going to be anywhere near as powerful as desktop Nvidia cards?

The performance is not that great, but you can already play Cyberpunk 2077 on a M2 Pro laptop using Game Porting Toolkit https://www.youtube.com/watch?v=sPJpkRmsceU Keep in mind, it has to do x64 -> ARM64 transition and translate Direct X into Metal. I would expect a native version to work better.

> Does this mean we can finally play Cyberpunk 2077 on a laptop?

My 2020 14" RTX 2060 laptop played it just fine, with decent settings (and copious DLSS). At 4K!

In the state I played it (earlier this year), it was great looking even without raytracing.

No. It's nowhere near as powerful as desktop Nvidia cards.

It's about as powerful as a PS5.

Do you mean at max raytracing settings? Alienware laptops should run Cyberpunk at 60, just not maxed out.

I'd prefer if there'd be higher RAM limits for each variant:

* ~~Ultra~~ Max probably has an adequate limit with "up to" 128GB

* Pro would be nicer to have "up to" (say) 64GB, rather than 'only' 36GB

* Base perhaps "up to" 32GB, rather than 24GB

There are some situations (running a VM or three with different OS(es)) where there really isn't a substitute for more bits.

Max is the one with the 128GB limit. We'll see what the limit is for the M3 Ultra, possibly 256GB.

IMO, if I'm doing that heavy of a workload, I'd rather SSH to a server instead of trying to do it on hardware optimized to work on a battery.

I just hope they stop selling 8GB variants entirely. All those do is make job security for tech support people.

Whats the point of re-parroting Apple PR and include: "the most advanced x" for each new generation something Apple?

I get it that they must say it but for the rest of us it's pretty obvious that they would say and it would make sense to develop something that will be less advanced than the stuff you already have.

You might make different trade offs but you surely would not do whatever the equivalent of de-growth be in computing.

Apple launched M2 MBPs only in January. I wonder if Qualcomm's Snapdragon X Elite (benchmarking now, but not in products until mid-2024) drove some of the timing on this launch: https://www.windowscentral.com/hardware/laptops/qualcomm-bri...

From what I’ve heard from folks connected with Apple, M3 was designed some time ago and TSMC delays on 3NM are why M2 happened at all. So, Apple is releasing M3 later than they’d have liked, and I expect we will see the M4 refresh faster than we would have otherwise.

ARM on Windows is going to be irrelevant for at least a decade likely forever.

There is a long-standing culture with the Mac that (a) backwards compatibility is minimal, (b) developers need to always keep their code current and (c) users should expect constant turnover in the platform. None of that exists on Windows. And when we last saw Windows on ARM the outcomes were that this wasn't going to change.

Which means it will never be a transition rather a fragmentation of Windows ecosystem into two equal parts. Making it far less compelling for everyone.

other way around i suspect. qualcomm knew this event was coming and wanted to front-run's apple's announcement.

especially as the macs ship much sooner...

M2 was late rather than M3 being early.

Oryon seems like it's going to be very competitive with the very underwhelming M3 design.

It's shiny but I think I'll stick to my upgradable RAM and two m2 slots + a RTX 3070ti and i7-12700H for a fraction of the price.

Something at Apple broke when they switched to their own silicon outside iPad and iPhone. They turned around the slow march toward parity on price:performance/RAM/storage and went screaming in the other direction. I really wanted to switch to mac when I was shopping for a new system earlier this year, but the prices were and remain ridiculous.

People have never bought Apple because of the hardware. They buy it because of the software, the aesthetics, the label, the support (It's easier to find an Apple repair shop than repair shops for any other brand anywhere in the world), and the integration with Apple's services. Comparing spec to spec is missing the point.

Also while this is niche, at the high end (M3 max + 128gb) you end up with a laptop that can literally just do things no Nvidia laptop can do right now due to the gobs of memory directly addressable by the GPU. I'm not saying 99.9% of people are going to take advantage of it, but it is a thing.

I tried to find a laptop with the specification you just mentioned that could last 20 hours on battery. Didn’t.

Would be interested in a link to a model if you have one handy!

You know they announced a price cut right?

Apple Magic Keyboard, Mouse and Trackpad are stil Lightning, I cannot believe it

The regular M3 chip is still software limited to a single external display...

I can't believe all their external monitors are still 60hz...

Love the improvement on the CPUs, but I was hoping for bigger iMacs and smaller Macbooks. I work at home on a 27" iMac, and on the road on a 12", 2 pound Macbook. We are years into the M generation of chips, and iMacs are still stuck at 24", and laptops are minimum 3 pounds and 13".

I feel similar to you, but I don't think we are going to see another high end iMac. Apple wants us to get Studios.

I purchased a fully maxed out M2 Max MacBook Pro when it was announced like 6 months ago. I knew it was going to be outdated eventually, but within 6 months, wow. Did not expect that.

i mean, if the machine does the job you purchased it for, is it really outdated? Unless the job it is doing is to be the best and latest, as a status symbol...

I use a maxed-out M1 Max and have 0 complaints. There was never a task where I felt the memory or CPU performance was not enough

You can feel slightly less bad. M2 Max was announced in January, 10 months ago.

I have an m1 max and I'm more than happy. Until local LLMs work really well on these chips I don't see a reason to upgrade

depends on how you define "outdated" i guess

I love Apple but they wasted a 30 minute midnight event in the UK to announce new Mac chips and a new colour of Mac. I think this could’ve been a press release and tech reviews hands on. I don’t think they’ll do this again but it wasn’t that good. Yeah filming was obv. But the event purpose I don’t think made sense for 30 minutes when it could’ve been a press release and tech reviewed hands on.

It was also showcasing the iPhone 15 Pro camera, as the entire event was filmed with that device

I have an 16GB M1 Pro and it's a great machine. I want to jump the shark on a M3 max. Feels like some computer that should last 4-5 years much like my m1. However I keep an ugly ass full AMD Asus ROG laptop running Linux for gaming.

I want to wait for the dust to settle on the Proton-like compatbility layer feature. I know crossover exists but past experiences lead me to believe is not as compatible as Proton. Back in 2021 I've tested various programs that worked well on Proton but not so well in Crossover. Hopefully the Metal enhancements fix it.

Hopefully, but at least from my (non-professional) perspective, Apple makes it difficult to make software for macOS

- no support for cross platform gfx apis (opengl, vulkan

- requires physical Mac to publish, no support for emulation

Ya I'm on an M1 Max with 64G of ram, and it's still great. I know the M2 and M3 now (for Max) have more memory bandwidth, which is attractive, but otherwise...meh.

On a related note, and on a personal level, my heart goes out to Johny Srouji and the Apple team in Herzliya. War is horrible, and I hope every reservist called up comes back to do magic at Apple.

It’s strange that Vision Pro suddenly no longer has the current chip. (But it’s possible that they announced it with the M2 and started working on putting in the M3 as soon as it became available internally.)

I think AVP will have M3. All of the new features screams AVP.

Interesting that the whole presentation was shot on an iPhone. I've been wondering when they'd make that leap.

What I really miss in my work issued M1 Pro isn't more power, it's the keyboard in those 2010s models.

I agree. And the longevity of them specifically for me. My M1 (MBP and MBA) keyboards shine after a few weeks' of activity whereas no shining on older MBPs with a decade of use – it reeks of a quality degradation.

110% I still use my Macbook Pro from 2011 and when using brand new Macs, I can't help my miss my home Macbook pro

Can we use more than one external display yet?

Seemingly no. So far it looks like (someone please correct me if I’m wrong): - M3, 1 external display - M3 Pro, 2 external displays - M3 Max, 4 external displays

This has to be a product differentiation decision at this point, right? Is there any serious technical limitation against more displays in the base chip, that Apple has not been able to solve in three generations?

Of course. Have been since M1 Max. My MacBook Pro M1 Max powers two external displays.

I'm sitting here on an M1 Max with two external screens ..

I so wish these beautiful beasts ran all the games for windows like Proton / Steam Deck :(

"No matter what your passions are in work or in life, there's a Macbook Pro that's perfect for you."

For the vast majority of Apple customers, Macbook Airs are plenty powerful. They're also lighter weight, and significantly cheaper.

M1 Max 16 GB struggles for my use cases, ram is way too low for how much I paid for it.

Then I look at the new line up and they are starting at 8GB still? And please don't BS me with "it's enough for most users", because this is a stupid argument, gonna tell you what, so is my macbook 2013!!!

An 8gb upgrade then costs me 300 euros? This is crazy, 8gb ddr5 memories at Apple volumes is 15$ expense at best.

The fact that Apple keeps gouging software engineers and professionals that will always take the upgraded model (as my employer and virtually all I know does) makes me say yet again: those models can sit on the shelf for me.

I'm not opening my wallet for such a greed.

Looks like the latest MBA is only 0.1 lbs less than the equivalent MBP at 3.4 lbs. Feels like they used to be lighter.

I miss the 27" iMac. It was (is) such a neat mac and it was even a reasonable deal considering it came with a nice screen.

I was hoping—if not expecting—to see a refresh of the 27” iMac; I was ready to buy one! It disappointed me they simply upgraded the current 24”.

The new 24” are lovely, but I will miss the real estate (still using a "Retina 5K, 27-inch, Late 2015” one).

The best feature to me is really the battery life and the quietness of the laptop. That's a game changer for me. I can work from anywhere comfortable, without burning my palms or legs and in silence.

I remember working on a PC laptop years ago and as soon as you pressed F5 the fans came on cranking for a couple of seconds, it was farcical.

There are Windows laptops with fantastic cooling setups. My old G14 can run fanless, and they've improved the cooling even more in subsequent generations.

The thing Macs really excel at atm is long, sustained workloads on battery... Other than gaming.

OK they got me, time to upgrade off my i9 MacBook Pro.

>Rendering speeds are now up to 2.5x faster than on the M1 family of chips.1 The CPU performance cores and efficiency cores are 30 percent and 50 percent faster than those in M1, respectively, and the Neural Engine is 60 percent faster than the Neural Engine in the M1 family of chips.

WTH, was Apple sandbagging on the M1?

5nm vs 3nm and three years of progress. Intel's e-cores have been seeing more generation over generation uplift compared to their p-cores so nothing about this strikes me as anything unexpected.

When the m1 came out details about the m3 were well known and that this was going to be a pretty big generation for performance improvements. I think the surprising thing was that m2 got pretty close to m3 in situations which were not as thermally constrained.

These are the new 3nm chips from TSMC, right?

Is this still ARMv8?

I think ARMv9 SKUs from others are just around the corner, and that's going to be a pain point for Apple.

https://developer.arm.com/documentation/102378/0201/Armv8-x-...

Apple implements all armv8.5 features and some 8.6 features.

The 9.0 competitors are at the same level as 8.5.

Cortex-X4 is at 9.2 which is not yet in any SoC.

The only difference between armv8.5 and armv9.0 is one single extension that is required on 9.0, namely SVE2 (scalable vector extensions 2).

As someone who prefers the MacBook Pro feature set (Pro Motion, namely), but doesn't need the best processor, I'd be fine with the 14 inch Pro, M3.

But that model doesn't come in Space Black. Well played, you son of a bitch. Well played.

They've mastered the art of the pricing ladder and it's fairly aggravating.

Well, you mention you mainly care for feature set. What kind of feature is body color that would affect functionality. In fact I feel any kind of scratch would be worse on space black as compared to gray or silver.

I’m sure I remember Apple using Myst to sell Performa Macs in 1995. Some things never change.

So they remove MacBook 13 inch and replace with MacBook 14 inch as base. Many people will see the $500 off. Good selling strategy.

Btw, I didn't see the comparison with M2 chip. Could anyone give me the reference for that? Thank you.

They had multiple slides for M1/M2/M3 in the video. M3 is around 15% faster than M2 in most benchmark (according to Apple).

I just felt the continued existence of the 13" MacBook Pro was just confusing because it was very arguably worse than the 13" MacBook Air even without considering the price.

> And, a new media engine now includes support for AV1 decode, providing more efficient and high-quality video experiences from streaming services.

It's taken years but AV1 has ended up everywhere.

Sadly, only a handful of GPUs out there have hardware AV1 encoding chips. If I read the product page correctly, that new M3 architecture still cannot encode AV1 from hardware.

Apple's biggest problem may be how good the M1 laptops were. I have a 14' M1 Max and there's no task in my workflow that's slow enough that I feel any real need to upgrade.

It's frustrating that they don't refresh the whole line at once. I'm in the market for a new mac mini but now I feel forced to hold off until it gets its refresh.

I got a new M2 Pro 16" just a few weeks ago. I was pissed when I discovered that a new M3 Pro was coming out. Reading all the comments here, I feel a bit better.

You missed out on black though.

"Graphics-intensive games like Myst have incredibly realistic lighting, shadows, and reflections, thanks to the next-generation GPU of M3"

This is the funniest thing I have ever seen in an Apple press release. Myst came out in 1993, and the "graphics-intensive" version of Myst that they are demonstrating is clearly realMyst which came out in February 2014, just a few months shy of a decade ago.

Meanwhile, my wife is happily playing Baldur's Gate 3 on her M1 Mac. Could they really not get the sign-off from any other studio but Cyan on this? No shade against Myst, it is one of the best games ever produced, but this is not how you do a graphics demo...

----

EDIT: Just now learning that there is yet another Myst remake from 2021 made in Unreal. It turns out I have played this version, but I did it on a Meta Quest 2, so I thought I was just playing a VR port of realMyst because it looked equally bad.

The realMyst from 2014 was actually realMyst: Masterpiece Edition, the Unity engine remake of the original realMyst from 2000 that used Cyan's own 3D engine (later used in Uru and Myst 5).

So Myst has to be one of the most-remade and ported games in the industry: five major iterations just on PC.

"the most advanced chips for a personal computer", says Apple. This is not an objective statement.

Name the more advanced chip then... Intel and AMD fall short.

Well, if you consider the entire package rather than CPU alone, I can see it. Still a misleading claim, but if stretch the definitions of “advanced” and “chip” you can do it.

What is the objective definition of "advanced"?

> This is not an objective statement.

Rick and Morty did it before.

I have found some deals on M1 Max Macbook Pros with 10-core CPU, 32-core GPU, 64GB RAM and 2TB storage. Does it make sense to go with those or with base M3 Pro with 11-core CPU, 14-core GPU, 18GB RAM 512GB storage which is the same price?

FWIW I dev on an M1 Max MacBook Pro and it’s the best computer I’ve ever owned.

Maybe the best thing about these M3 chips is that they’re making those older machines more affordable.

Fwiw, I have a 16" M2 Pro and had a close call with losing it to water damage, so I picked up a M2 Air and it is FASTER than the Pro for most of my jobs (electron/Xcode dev). The Air is so much lighter that I much prefer it to the Pro. The Pro now lives at my desk via TB, and the Air in my backpack for pretty much everything else. Weirdly, when I look at GeekBench browser I see that the M2 Pro should be faster, but at the time I've got screenshots of it being a Pro @ 2.7 GHz and Air @ 3.5 GHz... and I just checked again right now (at my desk) and the Pro is running at 3.29 GHz, so the dynamic CPU scaling must have had an effect. Yet, the actual work doesn't lie, and everything feels super snappy on the Air. Truth to YMMV.

No, keep the M1 Max. M1 Max has bigger RAM & better GPU compared to M3 Pro.

"The industry’s first 3-nanometer chips for a personal computer debut a next-generation GPU architecture and deliver dramatic performance improvements, a faster CPU and Neural Engine, and support for more unified memory"

This reads horribly in my opinion

That's Apple's hyping as usual. Apparently the improvements are not that "dramatic" when you look at numbers, at least not any more than M1->M2.

Finally time to upgrade my 2017 i7. For over 1 year the battery meter has said "Service Recommended" and in the last 2-3 months multiple key caps are literally falling off and regularly get jammed while I try to type.

I figure the 16" with M3 Max and 64 GB is a good machine that will last me 5-6 years but I'm not sure if I should pony up for the 2TB storage upgrade.

I just pray this revision won't have some unexpected problems. I'm actually glad it seems to only be a spec bump so I won't be taking on any new tech. Hopefully any issues have been shaken out from the M1 and M2 releases.

It's harder than ever to be excited to spend money on an upgrade, despite the overall machine being better than ever. RAM and SSD upgrades are comically expensive in Canada, maybe the former of which being slightly more justifiable. $500 for the ram upgrade $750 to ugrade from 512gb to 2tb, so really I'd be paying something like $1000 on just basic storage. It's kind of insulting.

Well, thats annoying. I just bought a M2 Max 16" MBP in mid September.

I wouldn't have expected a new revision in the same year. I also swear apple used to have a policy for this situation, but I can't find it now.

Always check this page before purchasing Apple hw https://buyersguide.macrumors.com/

Maximizing margins for the M3 Pro which makes up a large majority of MBP sales. The transistor count reduction adds up over thousands of 3nm wafers. They are probably getting twice as many M3 Pro chips per wafer over M2 Pro. Nudging power users to the M3 Max doesn't hurt either.

Personally, I don't mind trading 2 P-cores for 2 E-cores but the memory nerf annoys me, and I suspect the cache has been nerfed as well.

I mind it to the point where I cancelled my order (upgrading from an M1 Pro). I'd do the mid-range Max but I imagine the battery life would suffer too much on the 14".

It's been pointed that that marketing for the M3 Pro is the same as it was for the M2 Pro earlier this year - up to 20% faster compute than the M1 Pro. As is the wireless web battery life.

Sure this will be variable based on the workload, but on average, this is a barely there upgrade.

The M4 will likely flip things back due to the yield issue, and there I suspect we'll see compute will improve by 40%+. I also suspect we'll see that upgrade sooner than later (next summer?) as Apple likely wants to get off these chips ASAP.

I’ve been wanting to upgrade my 2018 Air but I’m not sure if I’m ready to take the plunge on this new architecture and what I may not be able to do with it… the intel Mac’s keep getting cheaper, and more tempting… has anyone found they couldn’t do something important or even fun on the new M chip? Clearly you can’t run Windows as a virtual machine, which I may seriously want to do for Visual Studio if I get a much more capable Mac.

The M1 bar here shouldn't be at the middle? https://www.apple.com/newsroom/images/2023/10/Apple-unveils-...

No, that would be 100% if it was

That looks impressive, but why they show animations at so low fps as a demonstration of performance?

Surely they should play it at 60 or more fps...

Not sure exactly, but the whole thing was "shot on iPhone" so maybe related

M3 should not be in the macbook pro line. it only support one external display, should this laptop be called 'pro'

What does the number of external displays have to do with the laptop being "pro", whatever that means? My Air drives a single 43" screen at home, I literally can't fit more screens on my desk. At the office anyone can grab a screen (they're nice, 4K and around 27-30" I think), some people use them, some don't (e.g. people with 16" MacBooks).

Apple should develop framework motherboards. Also their ridiculous upselling practices are in the full swing again.

I really hope Apple doesn’t do a yearly cadence with new chips. From all the comments I’m reading here there’s a lot of confusion and misunderstandings about why new chips that seem to underperform or lackluster performance. Maybe Apple should just do every other year for chip upgrades.

Wonder if the unified memory (up to 128GB) makes it much better for machine learning / AI things? On a Windows laptop you get maximum a Nvidia 4090 and with 18GB of RAM for the GPU.

Do I miss something? I wished I had more insight on Apple architecture support from the AI frameworks out there.

> Graphics-intensive games like Myst have incredibly realistic lighting, shadows, and reflections, thanks to the next-generation GPU of M3.

I thought this was funny considering the original Myst can run on a modern toaster. I'm guessing they're referring to a new Myst re-release?

Yeah, it's pretty pathetic they keep bringing up Myst. I definitely highlighted it as a laughing point when summarizing it to others.

My favorite part was when Tim Apple teased the new M3 parmesan cheese grater.

https://i.postimg.cc/TYpx7SVf/Screen-Shot-2023-10-30-at-8-48...

On the power consumption vs performance ratios, every chart showed the M3 cranking at higher overall power usage than the M1 when under load.

One of my favorite things about my M1 is the battery life. I wonder how the actual battery life will compare for the different market segments.

You can always turn on low power mode. I do this for extra long days.

Seems like apple's software is lagging far behind their hardware. To this day I can't reliably use Facetime for a call. Often times its ringing on my mac but the other use gets nothing. Maybe take some time to fix your software as well, apple?

I'm waiting for the reviews until making a decision -- unfortunately I have a corporate laptop M1 MBP so can't use a personal laptop for work

As usual Apple will charge arm+leg for 32gb/1tb config -- and I don't need the graphics really

I want to see low level instructions baked in to allow reliable, performant and efficient real-time emulation of win/x86 software, especially games. As in, I don't care about what chip is inside, I want my game library to just work.

Just wait for M3 Ultra which in past generations is literally just 2 Maxs. Imagine 256 of GB unified RAM (oh and I guess the 32-core CPU and 80-core GPU, though those aren't as impressive). Really would love to know how much that is going to cost

The ultras don't make a lot of sense for most professionals when, at that price point a threadripper and a 4090 would still eat it for lunch.

I have a ~7 year old 27" iMac. It wasn't that expensive at the time and the screen is still way better than a lot of what is on the market today. Now, it is only available in a 24" version.

Is time moving backward?

I guess the high-end iMacs and iMac Pros (remember the iMac Pro? 10 core Xeons upgradable to 512 GB of RAM!) just never sold very well, so they canned the idea of selling iMacs to pro users and instead chose to focus on the lower-end market that was previously served by the 21.5".

So it seems to me like the iMac has very much been repositioned as a stylish, basic desktop computer for the home and general office tasks now. No M3 Pro/Max, no bigger monitors, limited I/O, etc. I personally don't really like it because I have labs full of decked-out 27" iMacs that are now kind of left without a clear upgrade path, but I understand what they're doing.

Have you seen in person the new iMac? The performance leap from that is insane. The 4,5K retina screen also fits as much even if not more than your 27-inch

How did Apple become so successful making high-end CPU chips on their own so quickly?

Bring in people with expertise, buy outside expertise where you can, focus on the product you actually want instead of the product that will appeal to every possible use case for every possible customer that the chip marketing team can imagine, don't sweat where you can put the margin line between chip and system integrator, because they are both you.

I suspect any $2,600,000,000,000 company could pull it off, if they moved first.

Apple acquired PA Semi in 2008 then Intrinsity in 2010. They have been using their own ARM cores since 2012 with the A6 SoC

https://en.wikipedia.org/wiki/P.A._Semi

https://en.wikipedia.org/wiki/Intrinsity

https://en.wikipedia.org/wiki/Apple_A6

They've been designing the A-series chips in iPhone and iPad since the A4 in 2010.

20 years ago, Palo Alto Semiconductor was founded, and 15 years ago, Apple bought them in order to start building their own CPUs.

The iPhone chip was seriously competitive with laptops 2-3 years before the M1, it's just that due to the software environment few people noticed what was going on.

The real question is why Apple left it so long. They clearly wanted the first generation to be a clear success, but they could have probably pulled this off faster had they wanted to.

It wasn’t that quickly if you consider the years of iPhone chips leading up to it.

They've been doing this for years. It was just mobile only at first.

So quickly 10 years in the making.

Not quickly. People have been predicting this since the PA Semi purchase.

They first spent 10 years making their own mobile chips.

Probably billions of dollars in R&D and extremely clear executive guidance?

What is "quickly" for you?

They don't "make" them, they design it.

I don't think we're comparing apples to apples here (excuse the pun). These chips have something like 10% of the instructions that a typical x86 chip does. Once the big CPU players start producing the same kind of chips, I greatly expect Apple's power to performance advantage to drop significantly, if not be overtaken by the likes of AMD, etc.

Is it me, or does the "space black" not look that dark? I have a Macbook Air in "midnight black" or whatever it's called, and it seems darker than this hue, at least from watching the video.

Black MacBook

Some folks are going to lose their marbles that Apple released a black MacBook Pro.

> The CPU performance cores and efficiency cores are 30 percent and 50 percent faster than those in M1

...or 15% and 30% faster than M2 according to their graph further down, so it's basically another M2-level upgrade.

Why do some of the chips (m3 pro only?) have 18g and 36g models, vs more traditional 16g and 32g? Can't find any explanation for it, just a lot of reposting of the specs in various sites.

See this thread https://news.ycombinator.com/item?id=38078282 seems to be based on available LPDDR5X chip sizes?

Also curious about this.

> Graphics-intensive games like Myst have incredibly realistic lighting, shadows, and reflections, thanks to the next-generation GPU of M3.

I'm just out of words for this.

I think it's quite clever actually, because although it's sort of a laughable technical feat compared to other options for gaming, many of the people who both have the cash to spend on the higher-end mac AND who don't pursue any more modern games on other platforms probably would have been original Myst players

Selling a tick for a tock.

> "Games like Myst have incredibly realistic lighting, shadows, and reflections."

So we've finally come full circle. Next step, bring the old CEO back in an advising capacity.

Just as Qualcomm shows it's chips are getting better, Apple pounces! The layout looks very different from m2, it must have been delayed. Does anyone know if it's using HBM3?

Nah, it’s LPDDR5X

All the videos on that page are so slow and choppy on my 2019 MacBook Pro 16" in Safari.

If not intentional, their marketing should do it every time they want customers to upgrade.

Stockholm Syndrome?

The new M3 base model, along with only supporting one external display also loses a Thunderbolt port. A new shell / chassis just to remove a port. Courage!

The M3 base model replaces a computer with two thunderbolt ports. It now has MagSafe, 2 thunderbolt ports, HDMI, and SDXC.

How is that a downgrade?

Other than the fact that I can’t upgrade the OS and increasingly vendors only support M-series chips there is nothing wrong with my Intel MacBook Pro.

If you are not running latest, your system might not be fully patched: https://arstechnica.com/gadgets/2022/10/apple-clarifies-secu...

So there are downsides to not upgrading.

Finally, meaningful amount of RAM in lowest option (which is the only one most shops usually have in stock) for the Pro - 36Gb.

What an underwhelming event especially since the M3 is worse than the M2 in some aspects.

I was hoping for some Mac Pro news or an iMac Pro or other such things.

Not buying an 8GB laptop in 2024, and I have zero intentions of paying the 300 euros markup to double it.

Those macbooks can stay on the shelf for me.

Tldr; Much better than M1, we don't want to compare it to M2 because that highlights it's only an incremental upgrade.

Letting Apple lead things in the market has gotten us to a very ugly space in personal computing. These are prohibitively expensive devices that for 95% of the use cases, a chromebook would be better-suited.

All those intel machines that were so impressive in 2019 are running like crap now. All going according to plan, I'm sure.

I manage a fleet of about 200 of these things. They are easy to manage, and oh-so-nice to look at, but boy, they are a huge waste of money.

These things should only be used by engineers and creatives. Yawn.

Right, that is what the "Pro" moniker in Macbook Pro means

Somehow Apple never has supply issues, meanwhile the good half of Lenovo’s ryzen 7040 series is not available in North America.

Which M3 configuration is the best price/performance option for an average developer? Is there an obvious best choice?

Literally every single Apple Silicon MacBook with 16 GB of RAM can easily run your IDE, one or two VMs and/or some containers, your messaging app of choice and your browser. That's more than enough for most dev jobs out there. If your job actually requires more RAM or CPU grunt, you'll probably know. SoC, RAM and storage upgrades are kind of expensive from Apple, so the base models are always the most cost effective.

Not that I can see. I spec’d this:

SoC: 11-core CPU, 14-core GPU, 16-core NE Memory: 18 GB Disk Space: 1 TB SSD Power: 70 Watt Price: $2199

I personally would go for the 36GB Ram on this, bump the processor to the 12/18 core set and go for 2TB drive but the price skyrockets with these additions.

FWIW, MBA M2 fully loaded is a very sound laptop.

For an average dev, I'd honestly skip the M3 and go buy a 15" MBA with 16GB RAM.

"Games like Myst have incredibly realistic lighting, shadows, and reflections"

Am I missing something here? Is it 1994?

There's an updated remake of Myst with a proper 3d engine and free movement etc. Even so that remake is a few years old now. It's a really odd choice when talking about cutting edge graphics.

Maybe Myst was the first game to get an update to adopt the new APIs?

Would be fun to see pgbench for M3 Max mbp

In the same way that Wikipedians rush to update the pages of recently dead celebrities, I'm surprised there isn't a similar rush to update pages like https://en.wikipedia.org/wiki/Apple_silicon around events like this.

We do. I was just at my day job. If someone else doesn't get to it this evening, I'll add something.

Will Maccy be as fast as dmenu on my old Thinkpad is what I need to know as a M1 Pro user

Can i just have fucking 3-4 monitors support on the bottom line pro model. ffs apple

So what about that Cinema4D rendering performance that showed 2x vs the M2 Max?

Was that GPU or CPU?

I suspect most of the side by side comparisons, including this one, were GPU.

I hear the complaints about incremental updates and not enough reason to upgrade. My opinion is that it doesn't need any updates as it's nearly perfect the way it is. What it needs is a 3 year release cycle instead of yearly. Less e-waste and more visible spec change.

M3 Max looks like the top tier GPU is a 40 series rival am I missing something?

Wonder by M10 if we have ChatGPT like AI built-into the chip.

And still no 27 inch iMacs. Strange decision, that’s a great size.

It seems that the 13” macbook pro (m1, non air) is gone for good?

Am i the only one who is frustrated with the mac monitor options these days? I don't know anyone who uses a 27" desktop monitor, much less a 24". What year is it? And $5k for a 32" is insulting.

So among all the Mac now, what's the best buy?

But it's not a PC, it's a mac.

What should I buy? Honestly it’s all confusing as hell. 8gb ram seems like too little but increasing it is $$$$.

You mean a Mac :P

Can they drive two monitors this time?

Base 14" M3 MBP comes with 8 GB RAM at $1599 USD. That's pretty stingy.

Am I the only one shaking my head that they kept comparing the M3 chips against M1 chips skipping M2?

Just like the last product release spent more time on huffing their own farts (the Apple Watch is the first of its kind carbon negative product), there is no real innovation or something to garner interest for me here.

no one does any heavy graphics work a mac anymore. especially for their showcases of rendering... its all nvidia GPU based for creation and rendered on farms anyway. Same with resolve.

Apple is so dumb.

If it doesn't run Windows, is it really a PC?

So MacBook Air chads win this round.

I'm looking for advice as I was planning to purchase the 14" M2 Pro MacBook, but I've noticed that the M3 Pro MacBook Pro has been recently released. I'm uncertain about which one to choose, and I'm seeking assistance in making a decision.

This is cool, where can I buy the chips and read their datasheets?

Pretty underwhelming a bit like the A16 release. They basically upped the frequency in A16 and that's how they got their single core performance improvement, not much of an IPC increase to speak of. Looks like the it will be the same story here given the performance numbers Apple gave. A16 also came with a power consumption increase so the 3nm process from TSMC is disappointing as well.

> Apple unveils M3, M3 Pro, and M3 Max, the most advanced chips for a PC

Nice! When can I order these most advanced PC chips for my next build? Do they have any kind of upgrade path for existing AMD/Intel users? I don't mind starting over too! Exciting times.

Reply on news.ycombinator.com