Apple frustrated most of us with mealy-mouthed comparisons and opaque speed benchmarks when it unveiled the iPhone 13 and its A15 Bionic chip. It seems the company could have made big claims, and the fact it didn’t may be a Mac story.
A spark in the dark
Think back to the announcement and you may recall Apple chose not to give us solid comparison data against the A14 Bionic chip, instead offering comparisons against a previous model of the device. So we had to try to guesstimate what the real performance improvements might be.
I have no idea why it chose to do so, given the strategic important of its processor designs. As is customary with an information gap of this kind, Apple’s critics quickly began sharing weird stories claiming the silicon design teams are suffering a brain drain.
Perhaps they are. Perhaps they aren’t.
But Apple has many engineers, so staff churn is to be expected. Misinformation grows in darkness, so it’s business as usual to my eyes.
What we’ve learned
Apple at launch called the A15 Bionic up to 50% faster than its competition. The four-core GPU in the iPhone 13 and iPhone 13 mini (five GPU cores on both Pro models) offers 40% (50% on Pro) better graphics performance, it said.
This wasn’t true. In fact, the truth is more impressive.
Fresh benchmarks from AnandTech show the A15 Bionic to be 62% faster than competing chips. While its processors seem to need more power, they are more energy efficient, with 32MB cache on the chip. This makes the processors more efficient again by keeping much of the routine operations on the processor, rather than driving these through memory.
The report also identified a 50% improvement in the L2 cache on the performance cores, and notes that the 12MB L2 cache on the chip matches that of the M1 processor on Macs.
What about Macs?
Given that Apple’s A-series processors are siblings to its M-series Mac chips, what’s good for one is good for both. And this suggests some of the iterative improvements we’re about to see appear in the M1X-powered MacBook Pro’s we expect Apple to introduce soon.
(While the potential introduction of these new pro Macs has been an intermittent rumor all year, soothsayers seem solidly behind the idea these will appear this month.)
Of course, translating the performance gains we’ve seen in Apple’s iPhone 13 chip into relative potential gains for the M-series Macs is deeply speculative, uncertain and cannot be relied upon. So, with that caveat, let’s speculate:
- If Apple has been able to unleash a 60% improvement in comparison to competing mobile processors, it should be able to tweak additional performance on the Mac variant thanks to the larger heat sink.
- Noting that the focus in this chip seems to be on energy efficiency, AnandTech observes a 17% power efficiency gain in the A15 compared to A14 chip. This should translate into a significant battery life gains on the new Macs.
The report also claims the A15 delivers a 28% performance uplift. Given the performance seen with the M1 Macs — best epitomized by Adobe Photoshop which runs at 1.5x the speed on them — it’s reasonable to anticipate similar gains on M1X Macs. This matters because Macs are used to get stuff done, and a 28% boost on the 50% (1.5x) increase Photoshop sees on an M1 means we can speculate (and this really is speculation) that Apple’s pro M1X Macs will deliver much, much greater performance than you get from the Intel systems they replace.
Announcing the M1 Macs, Apple claimed:
- 3D titles render 6.6x faster in Final Cut Pro.
- Project building is 3.6x faster in Xcode.
- Machine learning is 15x faster.
I suspect key Apple partners will already be working with these new Macs in their labs. We know Adobe was able to work closely with Apple to optimize Photoshop for the M1 chip. That work also bleeds into iPads, which already share the Mac chip.
AnandTech also says: “In the GPU side, Apple’s peak performance improvements are off the charts, with a combination of a new larger GPU, new architecture, and the larger system cache that helps both performance as well as efficiency.”
Mac users in pro markets will be curious to see whether this kind of graphics performance gain also comes to the next-generation M-series chips. If so, this translates into real productivity gains and opens the window to further creative opportunity.
Given the current M1 Macs provide 17-hours of battery life, the possibilities for mobile work are boundless. Perhaps we’re looking at a Mac you can use for days without a power supply.
Given Apple’s silicon teams are already designing processors to put into products in 2025, all of this bodes well for Apple’s platforms. “When we design our chips, which are like three or four years ahead of time, Craig and I are sitting in the same room defining what we want to deliver, and then we work hand in hand,” Johny Srouji, Apple’s senior vice president for hardware technologies, told Om Malik.
Now, I don’t know the extent to which any of these speculations will be met, but with a Mac event anticipated later this month, we won’t have long to find out.
NB: We’ve all gotten into the habit of calling the next-generation M-series chips “M1X,” but don’t be surprised if Apple calls them something else — M1+, for example.
Please follow me on Twitter, or join me in the AppleHolic’s bar & grill and Apple Discussions groups on MeWe.
Copyright © 2021 IDG Communications, Inc.