Apple’s Q2 2026 Results

It would be fair to say that controlling key technologies is quite important if there is not enough competition in them, and that makes Apple be at the mercy of the providers of those technologies. They see sufficient competition in the AI field now.

Remember when Steve Jobs bought all of the RAM? Fun times. Past times.

It's so weird to me that we talk about, even celebrate Apple's record-breaking quarters while they mistreat unionized employees and nickel and dime us at every turn

I feel like these should stop being celebrations and start being moments where we push Apple to start giving back to the people who made them so successful

> Remember when Steve Jobs bought all of the RAM?

I think that was flash memory (as in storage), and Tim Cook would’ve been COO then.

And that’s arguably exactly the strategy they’re still applying, which is why they can do the MacBook Neo.

But… I checked yesterday, and prices of 32 GiB sticks are up 490% in 365 days. Almost six times the price in just a year. There’s a point where even buying up an entire production run isn’t enough.

"Margins are now at 49.3%."

That's our money. Money they could have used to make products or services more affordable, or to cut back on advertising, or to invest in actual curation of their shitty stores.

"Either Apple is right and the rest of Big Tech will have lit hundreds of billions – perhaps trillions when all is said and done – of dollars on fire, or Apple is going to be in big trouble."

What trouble might that be, exactly? Worst case, they'll poach a bunch of people from failing companies like Anthropic and OpenAI and build their own model in a year or so. Or maybe buy one of the companies outright.

But why would they? Models are substitutable. That would be like buying a mining company to ensure they have enough quartz to make silicon wafers.

Perhaps in a few years, when hardware cost normalizes and it's more established how to train models, it makes sense for Apple to make their own, but what exactly are they risking by not doing it now?

@Plume Yeah, it’s really frustrating to see these record numbers alongside Apple’s lack of investment in the developer/customer experience. If they feel the need to charge subscriptions for keyboard shortcuts and expand advertising when they have these margins, are they going to go full Adobe if the numbers ever drop?

I don’t know enough to gauge what they’re risking. Are we sure this is more like Web engines (where having your own is important for integration) vs. Web search (where Apple was right that it doesn’t matter for them)? Is poaching people enough or if they wait too long would it end up like Bing, always putting them at a disadvantage?

There are currently at least five different Chinese open-weight models that are competitive with American SOTA models (GLM 5.1, Kimi K2.6, Deepseek 4, Qwen3.5/Qwen‑Next, MiMo‑V2.5 Pro). A lot of the information on how these models are trained is published publicly.

Unless something drastically changes, I don't see anything stopping Apple from making a competitive model if they are willing to invest the money needed for pre-training.

"if they wait too long would it end up like Bing"

I don't think Bing's abysmal search results are a product of them waiting too long. I think this is institutional incompetence.

I wish there was a new generation that covered Apple more critically than Jason “yeah the product got worse but just get used to it” Snell and co. I’m sure there is, maybe at Ars and so on, but it really feels past time that the most prominent voices on Apple coverage be different names than most of the ones shown above.

Why not also provide the number of units when they publish their quarterly results? And detail the part of the App Store in the services revenue?

I would also be interesting in knowing the number of the new iPhone Air units sold. And the new Vision Pro.

>> Remember when Steve Jobs bought all of the RAM?
>
> I think that was flash memory (as in storage), and Tim Cook would’ve been COO then.

Sören, you're correct, I was mistakenly thinking Jobs and DRAM when it was Cook and NAND.

Recent reports indicate Apple has planned ahead yet again (unsurprisingly) at a high premium, but it sure seems like they are still facing quite the supply challenge.

I wonder whether we'll look back on the Mac Studio M3 models as an incredible value, even after the M5 models are released. Perhaps they will maintain or even increase in their resale value, similar to used cars in recent years.

Aside, I started encountering application crashes on one of my tower PCs a couple of weeks back. DDR4 fail, one module with errors. The price for a pair of 32 GB modules, despite their age/speed, was going to be more than price of a Mac Mini. Yikes.

Visual inspection using a loupe revealed two 0603 surface mount tantalum capacitors lifting off of the board, with fractured, cold solder joints. A heat pipe was heating up the metal housing of some solid caps nearby on the main board, and those cans happen to be really close to the damaged caps on that specific RAM module.

Thank goodness I have the equipment to measure the caps were still good, and also have the soldering skills to fix the problem. 20 minutes not including the RAM tests, and less than a dollar supply cost. I wish everything was so repairable.

> I wonder whether we'll look back on the Mac Studio M3 models as an incredible value, even after the M5 models are released. Perhaps they will maintain or even increase in their resale value, similar to used cars in recent years.

Well, that didn't take long. Apple discontinued Mac Studio M3 configurations with more than 96 GB RAM, including the 256 GB configuration. Delivery estimates are now 9 to 10 weeks. No Mac Mini 32 GB or 64 GB configurations, either.