Note: As an Amazon Associate we earn from qualifying purchases.

My experience with Apple’s new M3 chips so far

My thoughts on Apple’s M3 chips: design updates performance leaps software support battery life cost and pro implications
Author
Affiliation
James Birkenau

TensorScience

Published

November 23, 2023

Introduction

Apple’s new M3 chips seem quite promising. As a developer, I was curious about Apple’s promise of increased efficiency and performance: lowering compile times dropping and speeding up related. Below is my experience with them.

Prices

Initial Impressions and Aesthetics

Photo of the new m3 macbook pro on a desk showcasing its design

When Apple announced the M3 chip, my tech enthusiast side couldn’t help but perk up. On paper, these chips are a marvel of engineering: they pack more performance per watt than anything we’ve seen before, and that’s a narrative I’ve followed with keen interest. My first brush with the M3 was through hands-on reviews and spec sheets. Let’s just say, the jump in core count and efficiency figures were impressive enough to make me seriously consider the environmental impact (and guilt) of discarding my still-nimble M1 MacBook.

Seeing the new lineup, I noticed the design hadn’t radically changed. It was somewhat reassuring to see the familiar sleekness—Apple’s hallmark—aesthetics that have, over time, almost become an extension of professional identity. It’s akin to a silent signal of sorts, reflecting a standard you uphold. That said, the darker hue variants are subtle; they whisper refinement rather than shout innovation.

The tactile experience of a new Apple product, though, is undeniably pleasant: the cold aluminum, the crisp screen, and the reassuringly firm keyboard offer a sensory experience that complements the internal horsepower well. I imagine that owning a machine infused with the M3’s prowess would make mundane tasks feel like a breeze and intensive tasks less daunting.

But beyond looks and the feel-good factor, any new chipset introduction ultimately probes practicality. For instance, the integration of more Thunderbolt ports speaks to a more pragmatic approach from Apple. It’s a gesture that acknowledges a world that isn’t quite ready to abandon cables, and in my work, juggling peripherals without a hub is a minor but valued improvement.

However, beneath this sheen, there are pain points. The soldered SSDs and RAM come to mind—hostage to a decidedly non-upgradable architecture. The lack of user-upgradable components is a thorn in the side for those of us who prize longevity and DIY fixes. It’s a reminder that for all its innovation, Apple’s grip on control remains as firm as ever.

In my head, I weigh these drawbacks against the benefits. The internal dialogue often circles back to whether the overall gains justify the investment and the ensuing electronic waste. With the performance leap, Apple’s pitch almost feels like it’s saying “Yes, without a doubt,” but my sustainability conscience murmurs otherwise.

Before diving headfirst into this M3 sea, I’ve had to remind myself of the other sections of the ecosystem. There’s more to consider, from the cost analysis to energy efficiency and software ecosystem dynamics. The lure of raw performance and crisp design is strong, but the discerning user in me knows it’s part of a broader constellation of factors that must align for it to make sense.

Still, there’s a not-so-small part of me that marvels at what Apple’s engineering teams have pulled off with the M3. It’s the kind of leap that tech dreams are made of, albeit a leap taken with a parachute of practical considerations firmly strapped on. The strides Apple makes are undeniable, and even with the strains of offer and limitations, it’s an exhilarating time to be part of the technology narrative.

Performance Jump from M1 to M3

Bar chart showing the performance comparison between m1 and m3 chips

Ever since the release of the M1 chip, I’ve been closely following the evolution of Apple’s silicon, curious about how much performance improvement we could realistically expect from one generation to the next. Now, with the M3 series hitting the shelves, I can’t help but indulge in dissecting the performance jumps we’re seeing, and they’re quite something to ponder.

To put it into perspective, I remember how my M1-powered Mac felt like a revelation with its balance of speed and efficiency—a testament to Apple’s design prowess. But with the M3, it’s not just a step; it’s more like a leap, especially when we consider the reported halving in compile times for heavy tasks like Xcode builds. This alone could justify an upgrade for someone like me, who values every saved minute on a project.

However, it’s not all about raw performance. Efficiency is another talking point that stands out with the M3 series. The idea that we can get such power gains without a proportional increase in energy demands is impressive, and it’s a reminder of Apple’s commitment to performance-per-Watt—a metric that’s becoming increasingly critical in our energy-conscious world.

Of course, we must recognize that not everyone is going to experience earth-shattering changes with the M3 upgrade. For the average user whose workload doesn’t push the boundaries of their M1, the M3 might seem like an extravagance. Plus, I’ve come across some early adopters noting that their M3 laptops seemingly devour battery life faster than expected, particularly while handling specific tasks. It’s a small wrinkle in an otherwise sterling fabric of improvements, indicating that software optimization may need to catch up to fully leverage the new hardware advancements.

Technical enthusiasts often highlight specific benchmarks to quantify generational leaps, but for me, and many others, it’s about the tangible day-to-day experience. Will it make my workflow snappier, and can it handle multiple heavy applications without the dreaded spinning beach ball? So far, indications suggest that the M3 will deliver on these fronts, and then some.

I won’t deny the allure of the M3’s promise, but there’s a practical voice reminding me that we’re tip-toeing into a territory where the gains might surpass the needs of many users. I’m left weighing the promise of the M3 against the existing capabilities of my current machine, which—let’s be honest—still feels eminently capable and far from obsolete.

As a part of a community that often debates the merits of upgrading technology, I find it important to mention that the decision to transition to the M3 should not be based on the excitement of a shiny new chip alone. Instead, it should be a carefully considered move, especially if you’re someone whose demands don’t necessarily justify the latest and greatest.

In any case, the M3 chips are an exciting development in Apple’s history of innovation, adding another chapter to their narrative of pushing the envelope in personal computing. If you’re contemplating the latest models, you might find Deciding between a Macbook M2 Air vs Macbook M3 Pro (2023) particularly enlightening. Even with some caveats, the M3’s performance jump is evidence of Apple’s ability to maintain an upward trajectory in the evolution of their chips, which is something any tech aficionado can appreciate.

Software Ecosystem and Compatibility

Screenshot of the macos interface highlighting new software updates

I’ve been navigating the evolving landscape of Apple’s software ecosystem since the transition to ARM architecture began, and my journey has been a mixed bag. On one hand, I’ve watched macOS gracefully embrace its silicon siblings with ever-increasing synergy and optimization. Tools like Rosetta have mostly bridged the divide, allowing me to run x86 software without major hitches. It’s impressive to see Apple pulling off such a shift, where even Microsoft’s legacy support pales in comparison. The ARM-native versions of software like Xcode 1 and popular development apps have been nothing short of a revelation in terms of performance.

But there’s a catch. Key applications in my workflow that haven’t made the leap to ARM still suffer performance penalties. It feels like the early days of PowerPC all over again, where certain tools lag behind amid a promising sea of innovation. Sure, virtualization is an option, but anyone who’s tried emulating an x86 environment on M-series chips knows it’s far from a workaround—it’s an exercise in patience, particularly when dealing with hefty development environments or complex graphics software that isn’t yet optimized 2. And, let’s not forget, the lack of Boot Camp means dual-booting into a full-fledged Windows environment isn’t on the cards, further compounding compatibility challenges.

As for the open-source domain, it’s a mixed bag. Major projects often receive timely ARM updates, thanks in no small part to the dedicated community on platforms like GitHub 3. However, smaller projects can get left in the dust, lacking the resources or incentive to match Apple’s pace. My trepidation stems from the worry that we might see a shrinking in the diversity and vibrancy of the software scene as the demands of keeping up with Apple’s architecture shifts may cull the herd.

What’s more puzzling to me is Apple’s stance on external compatibility. There’s a part of me that wishes for a socketed version of these ARM chips to be available for hobbyist tinkering or OEM integration. Despite knowing it’s a distant dream, given Apple’s tendency for close-guarded integration, it’s somewhat disheartening. Imagine an open-source community thriving around modified versions of macOS or M-series chips, much like the vibrant scene around Linux for ARM 4. The potential for innovation and customization is staggering, but alas, currently out of reach.

On the bright side, tools like Asahi Linux 5 offer a glimmer of hope by bringing a full Linux experience to M1 machines, hinting at the potential for the M-series line. However, for this ecosystem to truly blossom, it has to go hand-in-hand with the blessing—or at least the tacit acceptance—of Apple, which isn’t guaranteed.

In summary, Apple is pulling off a mostly successful transition that has its share of hurdles. As someone deeply embedded in the tech realm, I’m both excited and cautious. Navigating this new territory requires a bit of give and take, patience, and a readiness to adapt. But with each software update and community-driven breakthrough, it’s becoming clear that the M-series chips, for all their idiosyncrasies, are charting a new course for computing performance that’s hard to ignore.


Battery Life and Energy Efficiency

Battery icon with graphics indicating the improved battery life

Battery life has always been a cornerstone of the mobile computing experience, and with the advent of Apple’s M3 chips, expectations were sky-high. As I dug into the specs and user experiences surrounding the M3, I couldn’t help but be impressed with the efficiency metrics being touted, especially considering the processing power these chips pack. It’s clear that Apple is making substantial headway in performance per watt, a metric that’s increasingly crucial as we strive for sustainability and longer periods away from the tether of a power outlet.

Exploring the M3, I’ve found that while the peak power usage in notebooks seems to hover around the same levels as the M1 and M2—around 20-25 Watts—what’s extraordinary is the idle efficiency. The ability to sip just 10% of that power at rest means that for everyday tasks, be it browsing or document editing, the M3 is an endurance athlete. The Pro variants, predictably, lean more towards higher consumption, but they bring a proportional performance increase to justify that usage. A practical user like myself finds comfort in knowing that the battery isn’t just draining away on standby, and there’s a certain peace of mind knowing that I can work through long flights or coffee shop sessions without a desperate search for power.

However, it’s not a perfect picture. For those pushing their machines with more demanding tasks, such as video editing or heavy code compilations, the M3’s power draw becomes more apparent. It seems the “Pro” in M3 Pro might as well stand for “Power-thirsty” under load. It’s the trade-off for that blistering speed, and while it’s a justified one for many users, it does mean that for really intensive work, a charger should be kept at arm’s reach.

There also seems to be an ongoing conversation in the tech community about cooling systems, which are crucial for such high-performance cores. Socket SP5 supports up to a whopping 700W power draw, but that’s serious heat to dissipate from a relatively small area, no small feat of engineering1. This thermal constraint can influence power availability and ultimately impact battery longevity during high-performance scenarios.

Then we have the rabbit hole of AV1 hardware decoding built into the M3 chips. This should, in theory, increase battery life during video streaming, as more energy-efficient codecs mean less processing strain2. But real-world user experiences vary, with some reporting significant power savings, while others notice little difference compared to older codecs like h264.

To sum up, the M3’s battery life and energy efficiency are, by all accounts, impressive. The dedication to a low idle power draw is particularly commendable, and advances like AV1 decoding suggest a forward-thinking approach to efficiency. Still, it’s important to acknowledge the heat management and increased power draw under heavy loads. While average users will be more than satisfied, power users should stay cognizant of their use case and proximity to power sources, thus maintaining a sweet spot between the M3’s enhanced capabilities and its power-hungry lineage.

Cost Analysis and Upgradability

Infographic depicting the cost breakdown of upgrading from older models to m3

When I consider the cost implications and upgradability of Apple’s latest M3 chips, I’m left with a mixture of admiration and a touch of skepticism. The performance gains, as touted by Apple, are nothing short of impressive, and for professionals whose work hinges on processing speed and multitasking capabilities, these upgrades can be a game-changer. I have a soft spot for the technological leaps we’re witnessing; it’s hard not to appreciate the sheer engineering marvel that is an M3 chip with its densely packed transistors and optimized power efficiency.

From a practical standpoint, though, the cost of entry is steep. An M3 MacBook Pro can set you back a considerable sum, especially when you look at configurations that push into 32GB of RAM or more. It’s significant enough that I have to weigh the benefits against the financial outlay carefully. For developers, the promise of shortened build times and more responsive virtual environments could very well justify the investment. But beyond the professional sphere, the decision becomes murkier.

On the plus side, the resale value of Apple products remains robust. If you’re someone who keeps their machine in good nick, you can recoup a meaningful portion of the initial cost when it’s time to upgrade. This aspect takes some of the sting out of the upfront expense. I can chalk it up as a partially redeemable investment, which is a relief in a market where electronics can depreciate astonishingly quickly.

The flip side is the upgradability issue. Apple’s move towards soldering components like RAM and SSDs to the motherboard means that what you buy is what you’re stuck with—no mid-cycle upgrades to extend the life of your machine. For those who prioritize the ability to upgrade hardware, consider building your own deep learning machine as it offers the flexibility to swap out components as technology advances. For me, this is a concern. I value the flexibility to upgrade as needs evolve, and Apple’s closed architecture feels like a constraint on the longevity and adaptability of its devices.

The trade-off here is of course the famed Apple ecosystem synergy—tight integration of hardware and software that provides a highly optimized experience. But, should a single component fail outside of warranty, it might render an otherwise functional machine effectively useless. A risk mitigation strategy could involve opting for AppleCare, which, while adding to the total cost, provides a modicum of peace of mind.

So when we talk about cost, it’s not just the sticker price. It’s the additional expenses and the potential for future costs due to lack of repairability and upgradability. You have to consider the total cost of ownership, which includes potential repairs, resale value, and of course, the value of the productivity gains you’ll hopefully achieve with the more powerful hardware.

Despite these concerns, the allure of the M3 can’t be denied. I can see scenarios across industries where this chip’s power efficiency and processing capabilities could mean significant efficiency gains. The M3 isn’t just an incremental update; it’s a symbol of a tide shift in computing, where mobile silicon is rivaling, and sometimes surpassing, desktop-class processors. If the majority of your workload can leverage the M3’s strengths, then the investment might just be a sensible one, despite the economic and ecological qualms one might have regarding upgradability.

Implications for Developers and Pro Users

Illustration of a developer at work with multiple screens displaying code and the m3 chip

As a developer, the shift from M1 to M3 chips by Apple has been intriguing, to say the least. The evolution in performance with the introduction of these newer chips has provided a somewhat compelling reason to consider an upgrade. In my experience, compilation times for large projects can be a real drain on productivity. With the benchmarks suggesting a sizeable jump in performance, especially on multi-core tasks, the M3 chip could save valuable time, especially for those of us dealing with Xcode or similarly demanding development environments.

But, let’s not trumpet the triumphs without considering the caveats. One drawback of this constant march forward is the feeling that hardware becomes obsolete at a more rapid clip than in the past. This is coupled with a non-trivial price tag for top-tier models with larger RAM and storage.

That said, I’m pleasantly surprised by the evolution of Apple’s Neural Engine on the chips. With machine learning and on-device inference becoming a significant part of software development, the doubling of compute capability on A-series chips is noteworthy. This boost hasn’t made it into the M-series chips just yet, which is somewhat of a letdown, but for iOS developers, the A-series power is a welcome advantage.

Another significant point for developers to consider is virtualization. Running different operating systems through virtual machines has been a thorny issue with the M-series chips, given the previous x86 architecture compatibility we enjoyed. However, tools like 86box offer a lifeline, though these are not seamlessly integrated solutions and might require additional setup and adjustments. Project repos like Apple Silicon for Docker give insights into the ongoing efforts for better software support in this domain.

Venturing into the realm of software support, Apple’s migration has been far from smooth sailing. While Rosetta aims to bridge the gap, it’s a stopgap at best. As someone who values a wide selection of tools and the flexibility to change my workflow, it’s troubling to witness the shrinking of software support with ARM transition.

For pro users, the energy efficiency is impressive, with reduced costs per core and extended battery life over their predecessors. The notion that greater power demands greater energy is dispelled somewhat with the M-series, making the use of these chips in a broader range of environments plausible.

With socket power, I’ve noted the lofty peaks that can be reached with top-tier CPUs like the Threadripper, but again, it’s a balancing act between pure power and system requirements. In the case of the M-chips, the combination of a relatively modest power draw with high performance is desirable, especially when considering deployment in data-sensitive or power-constrained environments.

In wrapping up, while the transition has had its pain points, the future looks bright if you’re willing to embrace the new architecture and its growing ecosystem. Despite the high cost of entry, the performance gains, improved efficiency, and evolving software compatibility suggest a net positive outcome for developers and pro users embedded in or transitioning to the Apple ecosystem. The trade-offs require careful consideration, but if your workflow can leap the hurdles, the M3 series chips are poised to propel productivity forward. However, it wouldn’t be imprudent to hold onto the M1 a bit longer, especially for those who aren’t hitting the performance ceiling just yet.