Less than a month from now, Apple is expected to officially unveil its new A9 chip. This will be the ninth A-Series processor including the original A4, which powered the first iPad, iPhone 4, fourth-generation iPod touch, and second-generation Apple TV. It’s hard to overstate the importance of the A-series chips to Apple’s devices, as they’ve helped the company to achieve everything from major processing leaps to impressive power efficiency and — often taken for granted — guaranteed UI smoothness for every year’s newly-launched devices.
With the iPhone 6S just around the corner, we’ve started to receive tips purporting to reveal how much better the A9 will perform than the A8 processors found in the latest iPhones, iPad Air 2, and iPod touch. While we wouldn’t characterize the numbers we’ve seen as reliable, they led us to look back at the history of A-series chips, and consider what can reasonably be expected from the A9. Read on for our thoughts…
How Each Year’s A-Series Chips Have Progressed
Most of the time, Apple introduces two new A-series chips in a year — for instance, the A5 for iPhones and the A5X for iPads, which were followed by the A6 for iPhones and the A6X for iPads. Every year’s chips improve on the prior versions, with slightly different priorities for the iPad and iPhone. The iPhone version is designed to be more energy-efficient and at least a little less powerful, while the iPad version generally runs faster and has more power, offset somewhat by the need to drive a higher-resolution screen.
Processor: | Single-Core Score: | Multi-Core Score: |
---|---|---|
A4 | 211 | 211 |
A5 | 220 | 415 |
A5X | 267 | 503 |
A6 | 708 | 1272 |
A6X | 768 | 1399 |
A7 1.3 (iPhone) | 1354 | 2500 |
A7 1.4 (iPad) | 1467 | 2652 |
A8 | 1610 | 2890 |
A8X | 1808 | 4526 |
As the table above shows, each year’s upgrades vary considerably in performance jumps, but one thing’s constant: every new flagship iPhone’s A_ processor is a step better than the last flagship iPad’s A_X processor. The A6 offered huge performance jumps over the A5X, and the 64-bit A7 did the same relative to the 32-bit A6X. By contrast, the A8 respectably outperformed the faster iPad version of the A7, though there wasn’t an A7X to compare against.
What Will The A9 Be Like?
We’ve already seen photos of the A9 processor on an early iPhone 6S motherboard, and superficially, it doesn’t look much different than the A8. But under the dark gray exterior, a lot has changed under the hood. Apple’s chip partners have moved from a 20-nanometer process to 14/16-nanometer FinFET manufacturing, enabling the A9 to squeeze even more transistors into the same surface area. This will let the A9 offer greater performance and improved power efficiency at roughly the same size as the A8.
An unsubstantiated tip suggests that the A9 inside the iPhone 6S achieved a Geekbench single-core score of 1921 points, and a multi-core score of 4873 points. For single-core performance, this would be a roughly 19% jump over the A8, and a 6% improvement over the A8X, versus a nearly 69% multi-core jump over the A8 and almost 8% over the A8X. (Note that the A9’s purported single-core improvement would be nearly identical to the A7 to A8 jump, and around 4.4 times larger than the A7 to A8 multi-core improvement.)
These numbers could be completely fabricated, which is why we haven’t run them in a news story. But they are in line with changes we’ve seen from Apple each year, and as such, a reasonable basis for contemplating what’s likely to actually happen. Again, as the table above suggests, there has never been a flagship iPhone that scored lower on Geekbench’s single-core or multi-core tests than the prior year’s flagship iPad. If there was going to be a first year for that to happen, it could easily be this year, as the A8X’s multi-core score is dramatically higher than the A8’s, thanks to both a third processing core and a faster clock speed.
To beat the A8X in both categories, the A9 would need to evolve from a dual-core processor to at least a triple-core processor — achieving a multi-core score superior to the A8X with a dual-core A9 would be virtually impossible. So it’s interesting that the purported multi-core score of 4873 is right in line with what we’d expect from a triple-core A9 with a single-core score of 1921. However, the same tipster claims that the A9 will feature two 1.7GHz cores and two 1.2GHz cores. It’s possible that Apple’s design does use two faster cores and two slower ones to achieve a comparable top speed to a triple-core design, while switching between cores as necessary to conserve power. That’s the type of compelling energy-efficient innovation Apple would be proud to unveil.
What Does All Of This Mean?
It’s possible that the next iPhone’s chip will fall below last year’s flagship iPad in performance, but Apple’s chipmaking history suggests that’s unlikely. As a result, we’d expect to see the iPhone 6S sporting some major performance gains over the iPhone 6. The extra power could be used for even more impressive games, processing 4K video, and running apps that are even closer to OS X-class. If the A9 matches or comes close to the numbers above, it will be in the same league as 2009 iMacs, 2010 MacBook Pros and 2011 MacBook Airs, all of which are still viable computers. Yet the iPhone will be profoundly more power-efficient, and unlike any of those Macs, capable of fitting in your pocket without any cooling fans.
This also bodes well for the A9X, which has reportedly been under development alongside the A9. If Apple follows its normal pattern, the next iPad will be on par with entry-level 2010-2011 iMacs, and quite possibly a step better than the entry-level 2015 12″ Retina MacBook. The key question will be whether Apple aims to close the performance gap between its tablets and laptops, or continues to limit the next iPad’s performance in the name of conserving battery life.
More From This Author
Check out more of my editorials, How-To guides, and reviews for 9to5Mac here! I’ve covered a lot of different topics of interest to Mac, iPad, iPhone, iPod, Apple TV, and Apple Watch users.
FTC: We use income earning auto affiliate links. More.
i expect something as powerful as Samsung’s Exynos 7420 but with a more powerful GPU and much more frugal
i really cant wait to see how Apple will surprise us all this year
A9 much more powerful than Exynos.
Nice!If true I look forward to a force touch 12″ ipad
With all this.. I look forwatd to having more comprehensive and powerful apps for serious content creation on ios.
> That’s the type of compelling energy-efficient innovation Apple would be proud to unveil.
Except that’s pretty much what ARM’s big.LITTLE arch is, something many high end ARM Android devices have been doing for a few years.
Except that Android devices are power hungry, and last always less than iOS devices with smaller batteries…
Ok, that has nothing to do with what I said. The point I was making is that Apple is not the first and is not the innovative company to come up with 2+2 cores mix, ARM did this a few years ago with big.LITTER. It’s the same thing.
Android is inefficient, that is why it needs big powerful, power-hungry processors to do the same tasks and enjoy a similar environment as iOS.
You’re mixing hardware and software up, Android has nothing to do with big.LITTLE, it’s a hardware CPU platform feature from ARM. Android just happens to be the first software platform to use it. Apple won’t be the first “innovative” company to do this, it’s already been done.
It would be so amazing if the Apple TV has an A9X over clocked and without any efficiency constraints due to it always having a power source. 2GB of DDR4 RAM and an Apple designed video game controller for the plethora of games that will fill it’s App Store. Granted, it will probably have an A8X, which is sad because it should be getting the A9 since it is at the iPhone event. Apple has the opportunity to make that App Store be a compelling reason to purchase the new Apple TV, but that requires either an Apple designed controller, or the ability to connect via Bluetooth the PS4 controller. No one buys the cheap garbage MFi controllers, and hardly anyone even knows about them. They NEED to break out of the insane restraints they gave themselves with MFi, and let this thing become great for gaming, and I don’t mean casual gaming which belongs on mobile devices on tiny screens. A TV screen needs a tactile, great designed controller, and hardcore games. If they had this, AAA developers would be happy to launch games on its App Store.
oh my god , this is indeed exciting, putting a 2010 mac into ur pocket !! I love technology!!
The quad core design with 2 fast and 2 slow cores make some sense with the iOS 9 power optimizations. Enabling the Low Power Mode saves significant power but also significantly decreases performance. I run two quick tests using the Octane Benchmark on an iPhone 5S with iOS 9 latest public beta. Normal: 6015 vs LPM: 4150. The numbers are not conclusive but provide some insight. That’s 32% reduction in speed. If i’m not wrong, adding 2 extra cores is “easier” than adding additional steps for lower power consumption to an existing core.
On the other hand, Apple loves simplicity and optimizations rather than throwing in more core for just the shake of it. Extra cores can be a challenge in programming and Xcode needs to do all the job transparently for devs, which means Xcode needs to be bug free and the power management on hardware to follow along the same lines.
So the next iPhone will have as much single-core processing power as my Mac Pro 1,1 which is still one of my production machines, albeit with 8 cores and 32GB of memory. But still, amazing.
I could be wrong, but, I don’t think we will see 3-cores, as I would think that would generate too much heat…not even the new, fanless MacBook has more than 2 cores.
nevermind. (A8X has 3 cores, duh).