How Apple solved the RAM crisis before it even arrived

Over the decades, the evolution of computers has been easy to follow: They’ve become steadily more powerful to keep pace with the need for more advanced applications. We’ve seen computers equipped with faster processors, more storage, and, most of all, more RAM. This trend has accelerated even further recently, with companies pushing on-device AI capabilities into their devices, and Apple is no exception.

However, the increasing demand for RAM is having another effect on the market. While once plentiful and affordable, RAM is more expensive than ever, and it’s becoming harder to get, even at exorbitant prices. But instead of dealing with constraints by upping prices, Apple is showing the industry how its years-long effort to control every component inside its devices has given it a unique advantage over every other PC maker, evidence no clearer than the recent launch of the MacBook Neo.

The importance of RAM in today’s world

First, let’s take a closer look at why RAM or memory is so important. Essentially, RAM is a type of short-term memory with ultra-fast speeds, much faster than SSDs. When you open an app on your phone or laptop, it’s stored in RAM to ensure that the processor can handle it more quickly as you continue to use it.

That’s why, generally speaking, having more RAM means your devices can run faster, enabling more apps and tasks to operate at the same time without slowing down.

As such, RAM has also become extremely important for artificial intelligence. Since RAM is much faster than internal storage, devices also store AI parameters in RAM for quicker access. In other words, the more and faster the RAM, the greater the device’s ability to handle AI tasks locally.

That’s why AI companies have been buying up as much DDR5 RAM, the same that’s used in Apple devices, as they can for their servers. It’s also why we’ve seen consumer desktops and laptops gaining more RAM in recent years, including Apple, which only recently increased the base RAM for Macs to 16GB and the iPhone and iPad to 8GB.

More RAM isn’t a viable option anymore

For years, the strategy of adding more hardware resources to handle software demands was a simple solution to complexity. But as modern workloads have become increasingly memory-intensive in recent years, suppliers have been unable to keep up with demand. As a result, RAM has become much more expensive, and PC makers have been forced to shift their strategy.

The situation became so dire that Apple had to discontinue the highest-end Mac Studio with 512GB of RAM, while other versions with higher amounts of RAM are also facing significant shortages. And it’s hard not to conclude that the timing of the Mac Pro’s discontinuation wasn’t tied to the RAM crisis as well. The RAM crunch has also been affecting the 16GB Mac mini, which is currently sold out at the Apple Store and unavailable at Amazon. Higher-end models with more memory are severely constrained as well.

This situation would (and has) put tech companies in a bind. They need better hardware to meet the growing demand for AI tasks, but there simply isn’t any more RAM on the market. Fortunately for Apple, it had already found the solution years before.

Apple chips were made for a RAM-scarce world

In theory, there’s no way to simply replace RAM with something else. However, Apple has found the next best thing: a way to optimize both its hardware and software to run smoothly even on devices with the bare minimum of RAM.

When Apple introduced the first Mac with the M1 chip, it wasn’t a mere swap from Intel to its own chips, but also a radical change to how its computers work. That’s because Apple silicon chips are classified as a system on a chip, which means key components, including the CPU, GPU, Neural Engine, SSD, and even the RAM, are all on the same pool, sharing the same resources. And that makes a big difference in everyday use.

Thanks to the SoC’s unified memory, all components can quickly access the same data without having to copy it between different modules. This drastically reduces latency and improves performance when running resource-intensive tasks without needing loads of additional memory. At the same time, Apple leverages its hardware and software integration to further optimize memory usage. For instance, macOS features a very aggressive memory compression system that significantly reduces the amount of RAM used by inactive apps running in the background.

The chips also utilize something called “memory swapping.” Admittedly, this isn’t a term Apple invented, but nowhere else is it used so seamlessly thanks to the way Apple silicon chips are designed. If your Mac runs out of RAM, the system can temporarily store memory data directly on the SSD.

Because both the RAM and the SSD are connected to the same SoC, data transfer is nearly instantaneous, allowing the computer to run smoothly even when you’re using all available RAM. When it comes to AI, the built-in Neural Engine also helps macOS process machine learning parameters and tasks, freeing up RAM. Essentially, this makes a Mac with 8GB of RAM feel like it has twice that.

While pro users will still need more, which Apple happily sells, the MacBook Neo proves once and for all that for everyday users, 8GB of unified memory delivers much more than the same 8GB of RAM in a traditional PC laptop.

The ‘RAM tax’ is already here

The timing couldn’t be better. In March, Apple introduced its most affordable laptop ever, the $599 MacBook Neo, while the rest of the industry struggles to stay afloat due to the RAM crisis. 

Microsoft recently announced significant price increases for the Surface lineup, with some flagship models now costing $500 more than they did at launch due to increased costs for memory and other components. For instance, the 13-inch Surface Laptop now starts at $1,149 with 256GB of storage and 16GB of RAM. That’s $50 more expensive than the base 13-inch MacBook Air model with similar specs. Unlike the MacBook Neo, Microsoft doesn’t even offer an option with 8GB of RAM.

Foundry

Meta also recently raised the prices of its Quest headsets by up to $100. The company likewise blamed the “global surge in the price of critical components.”

Apple isn’t immune to price increases for components, of course. The M5 MacBook Air, which was introduced last month, is in fact more expensive than the M4 model, starting at $1,099 with 512GB of storage rather than $999 with 256GB of storage. Still, it’s unlikely Apple has any plans to change the MacBook Neo’s starting price of $599 anytime soon.

That, of course, is because Apple has established that an entry-level Mac can run just fine with only 8GB of RAM thanks to its system-on-chip architecture. And even in the higher-end models, Apple silicon chips are so efficient that Apple is under less pressure to keep raising baseline RAM—and prices—as aggressively as competitors.

Eugen Wegmann

Efficiency will become a growing priority

As expected, Apple’s competitors finally seem to be catching on. Last month, Microsoft pledged to improve the quality of Windows with a series of improvements that will make Windows faster and more reliable by scaling back AI and freeing up RAM for other parts of the system, which should definitely improve performance on less powerful devices. Whether by coincidence or not, the announcement came about a week after Apple unveiled the MacBook Neo.

Apple isn’t reacting to this shift. It’s already there. Not because Apple devices don’t benefit from more RAM, they absolutely do, but because they rely less on memory specs to deliver a consistent experience. It’s no longer just about how much RAM a device ships with, but how well it uses every gigabyte.

Apple might not have designed the M1 chip with a RAM shortage in mind, but all this technology is now proving more useful than ever. This gives Apple a significant advantage over its competitors, who are unable to achieve the same level of integration with separate parts.