This is why your iPhone 16e performs worse than the iPhone 16

The new Apple iPhone 16e has a lot going for it: an affordable price, the latest iOS features, and flagship-tier performance. Well, that last point comes with a little asterisk. The Apple A18 processor inside the iPhone 16e isn’t quite the same as the A18 in the iPhone 16 or 16 Plus, which itself is already a downgrade on the A18 Pro in the latest iPhone 16 Pro.

While sliding performance for more affordable products is very much the status quo, the fact that Apple brands the iPhone 16e as having an A18 processor might lead many to incorrectly assume they’re buying flagship performance on the cheap, which they certainly aren’t.

You see, Apple chucked out (or rather disabled) one of the GPU cores inside the A18, leaving the iPhone 16e with four rather than five graphics cores for gaming and a few other tasks. It did the same thing with the A15 Bionic inside the iPhone SE (2022) as well. As you’d expect, this results in a significant performance drop of some 20% in our benchmarks, which translates over to real-world games as well.

Incidentally, it also seems that the CPU is downclocked slightly. Running Geekbench 6 shows a 5% drop in multi-core CPU performance, but single-core remains virtually unchanged. That’s not going to impact the responsiveness of your apps, but clearly, the Apple A18 inside the iPhone 16e is not quite the top-tier piece of silicon as the one in the more expensive iPhone 16 and 16 Plus.

Apple isn’t alone in this practice. The Samsung Galaxy S24 FE also has a cutdown but clearly rebranded Exynos 2400e version of the Exynos 2400 processor that powered 2024’s global Galaxy S24 and S24 Plus models. Again, we clocked a roughly 5% win for the flagship model in CPU tasks and a slightly more significant 11% difference on the graphics front.

Google’s mid-range Pixel A-series also performs marginally worse than their flagship counterparts despite using the same chip. We’ve seen this with the Tensor G2 and G3, and expect the Tensor G4 in the new Google Pixel 9a to follow suit, owing to its use of a more affordable manufacturing package. However, Google’s affordable phone seems to perform closest to its flagship sibling out of these three brands.

But why are these brands even bothering with weaker versions of the same chip for their affordable phones? Couldn’t they just buy something else? Well, it’s all about costs, of course.

Despite best efforts, not all chips come off the manufacturing line exactly the same. This “silicon lottery” means that two chips of the same design might not be able to handle the same peak frequencies. For example, one CPU might be able to hit 3.5GHz while another stops functioning at 3.2GHz. Likewise, sometimes chips come off the line with individual cores that are less capable or sometimes don’t even function at all.

Chucking away the less-than-perfect chips would be hugely wasteful, greatly reducing the profitability of a silicon wafer. Vendors could target the lowest common denominator, but that’s not great either. Instead, chip “binning” is commonplace in laptop and PC parts; the most capable chips go into the highest-priced components, and inferior parts of the yield are sold off as more affordable products. This practice hasn’t been so common in mobile chipsets simply because there isn’t a wide range of products to put the weaker chips in; consumers expect consistent performance no matter which Galaxy S or iPhone flagship they buy.

However, the move to smaller and far more expensive manufacturing nodes, from 5nm a few years ago to 3nm today, has made low-yield wafers increasingly costly. This is likely why brands that design their own silicon (Apple, Google, and Samsung) are increasingly keen to re-use flagship chipset designs in their more affordable products. In addition, there’s now a cheaper seven-core Qualcomm Snapdragon 8 Elite (with one efficiency CPU core disabled, making it around 14% slower). Qualcomm calls this an optimization for specific use cases, but I think we all know it’s simply trying to waste as little silicon as possible.

There are other ways to cut processor costs, too. For instance, Google swaps to cheaper packaging technology — containing the interconnects, thermal dissipation features, and protective packaging around the silicon — when building Tensor for its A-series phones. This also explains differences in thermals and sustained performance, in addition to the silicon lottery effect. It also downgraded the modem paired to the processor in the Pixel 9a from an Exynos 5400 to an Exynos 5300. Apple has done something similar to the iPhone 16e, using its custom Apple C1 modem, which lacks mmWave 5G, rather than Qualcomm’s Snapdragon.

Still, that’s another small change to ensure less silicon goes to waste, as well as providing consumers with a virtually flagship chip at a more affordable price point.

It’s worth reiterating that these performance differences between affordable and flagship models are typically only minor, in the region of 5% – 10%. However, Apple’s decision to drop an entire GPU core in the iPhone 16e (and iPhone SE before it) is by far the most aggressive cut-back out of these three silicon designers. As such, marketing the iPhone 16e as featuring an A18 chip like the iPhone 16 and 16 Pro is much more disingenuous.

Still, this all works out as a win for everybody. Smartphone silicon can keep chasing the performance dragon at the high end while mainstream consumers pick up the tab with slightly inferior chips that still perform brilliantly for the tasks they need on a daily basis.

Perhaps most importantly, this mid-range silicon still contains the same image processing, machine learning, and security components as the high-performing models. This trend’s biggest boon has been that budget-conscious consumers can dabble in the same cutting-edge innovations as more expensive flagships just a few months later.

Contrast this to the start of the decade, where many midrange products were stuck with far inferior performance, let alone features, compared to their flagship counterparts. Today’s iPhone 16e, Galaxy S24 FE, and Pixel 9a might not top quite benchmark charts, but they’re really not far off and cost just a fraction of the top-tier models. I can see why they’re so popular.