Tech

iPhone 13: what is changing with the Apple A15 Bionic chip?

One of the main changes brought by the new generation of iPhone is the arrival of a new chip. The Apple A15 Bionic is said to be more efficient and less power hungry than any other system on the market. Here is what we know about her.

Along with its new iPhones, Apple unveiled its latest A15 Bionic chip at its September 14, 2021 conference. At the time of its presentation, Apple assured that the chip was more powerful and less energy intensive than last year’s model, while being ” 50% faster Than competitive systems (understand Qualcomm chips). But, as often, the interest of this announcement is hidden in the details.

There are two A15 chips

The iPhone 13 and 13 Pro may both embed this new in-house chip, the two mobiles are not quite equal. If we compare the two models on the Apple site we see that the iPhone 13 Pro have a ” new 5-core GPU »While the standard models must be satisfied with 4 cores on the graphics part.

The A15 chip therefore exists in two distinct versions, depending on whether you opt for an iPhone 13 or 13 Pro. This difference can potentially be explained by the fact that the Pro model includes exclusive features, such as “ProRes” video recording (for image specialists), which require more power. This allows Apple to assert that the A15 of the Pro models is “ fastest smartphone chip “, While that of the standard models is content to be” faster than the competition “.

This is not the first time that the same chip has arrived in two different configurations (it is even frequent on Android mobiles), but the limitation imposed here by Apple seems a little artificial. Unless the manufacturer has actually built two chips on two different production lines – but that seems unlikely for reasons of cost. Deactivating a heart can also be a way to save energy.

15 million transistors

Now that this distinction is made, we will refer in the rest of this article to the A15 chip as a whole, as the differences between the two versions are in fact minimal.

The characteristics of the A15 Bionic chip // Source: Apple

On paper, Apple’s new component doesn’t look all that different from last year’s model. The chip is still engraved in 5 nanometers and there are still 2 cores dedicated to performance and 4 cores dedicated to ” high energy efficiency “. This architecture, after all fairly standard in the IT world, bears the name “big.LITTLE”. It is nicknamed like this because it offers powerful “big” cores for very demanding tasks, as well as smaller energy-efficient “small” cores, dedicated to more everyday activities.

It is therefore not on the raw technical sheet that the A15 differs from the A14. When we look in more detail, however, the differences are visible. Apple’s new chip contains 15 million transistors compared to 11.8 on the A14. This should translate into a slight gain in raw performance, but Apple does not communicate much on this point. The reason ? The focus has been put elsewhere.

A chip cut for AI

The A15 chip is indeed shown above all as a very talented component in the processing of tasks related to artificial intelligence. Apple has spent a lot of time talking about improvements to its machine learning engine, the Neural Engine.

According to Apple, this dedicated chip is capable of performing “ 15.8 trillion operations per second “. This should make it easier to handle complex tasks like using augmented reality apps or on-the-fly photo quality optimization. Apple continues here the trajectory “IA” started since the iPhone X. The competition is also doing the same, since with its Pixel Neural Core, Google already relies heavily on artificial intelligence to improve the photo quality of its smartphones.

Apple therefore does not rely on the profligacy of power to sell its new chip (after all, the architecture and the fineness of the engraving are similar to the previous version). Where the component stands out is in energy and software optimization. The A15 consumes less than the previous generation chip, while being more efficient with complex tasks. The proof that everything does not always depend on the numbers.

Related Articles