Apple Announces the M1 For Macs: 5nm Process, more than 2x CPU & GPU Performance and much extensive Machine Learning Applications

Earlier this year, Apple announced that it would be transitioning from the traditional Intel chips to a custom Apple Silicon. Today, we do not have to refer to the chipset as such. Today, Apple marked the launch of its first in-house chipset, based on an ARM process. The Apple M1 was announced at the November event today, making it the big highlight of the entire program. There are many features to highlight from the entire event and we shall try to incorporate all of them in the article.

5nm Process

New 5nm Process

Ever since the iPhone 12 launch, the company has boasted its first-ever 5nm process. This became quite certain that the company would bring it to the laptop chipset as well. Since it is a mobile processor, Apple didn’t hesitate from bringing forward its M1 chip to be the first chipset for a mobile computer to be based on a 5nm process. This means that it would definitely deliver more power and consume less energy, having a good ratio.

Design & Internals

A unified design for the chipset

Instead of going for a traditional style where the CPU, GPU and memory units are separate, the M1 chip groups all of these together into one form factor. This saves space but raises the question of whether it would be easy to cool the thing. They do it in their cellphones but can they handle the heavy performance of a laptop. The chipset also has the T2 chips, crammed in there as well. They claim to have done this as both the GPU and CPU would be drawing memory from the same DRAMs. The CPUs are 8-core with 4 high-performance cores delivering full-throttle performance and 4 high-efficiency cores. These would give adequate performance but with a fraction of the power consumption. These would obviously be handling most of the mundane tasks or simple web browsing.

CPU & GPU Performance

The New CPU shows performance gains

Now, when Apple announced the Apple Silicon, the main idea behind it was to ensure maximum performance while drawing the minimum amount of power. This time around, in their announcement at least, the company’s focus was just this. Right off the bat, they claim that the Apple Silicon or the M1 chip would be 3.5 times as fast as a similar PC. It would do so while consuming just about 25 per cent of the power. Apple claims this shows a performance gain of about 3 times. This would be significant as we would be seeing this performance at a lower price point. Not to mention, Apple demonstrated how app integration would help them run apps better on the M1 chip as compared to the regular ones we have seen in the past.

Twice as much graphical performance

As for the GPU side of things. Apple claims that this would be the fastest and most advanced integrated GPU ever. It would perform 2 times as well as the competition but by only using 33 per cent of the power required. Additionally, the event stated that this could deliver up to about 2.9 teraflops of graphical power. It is significant given the chipset and the form factor it would be coming from.

Machine Learning

New Machine Learning applications

The whole point of a custom chipset and an in-house process was that it would let Apple integrate its system better. This is what the company does in its iPhones as well. Considering the performance we have seen from normal components in Apple devices such as iPhones and iPads, this does not seem like a bad idea. Apple would be using Machine Learning in so many applications. The M1 chip would be about 15 times faster at ML processes. This would allow the chipset to work about 11 trillion operations every second. Apple is going to use this to enhance the battery life which has seen serious gains in both the MacBooks announced. Additionally, this would be used in the camera to give better result in facetime. Apps like Final Cut would benefit from this as well. People would, for the first time, be able to playback 8K raw footage on MacBook Pro 13 without any frame skipping. If this is true, this would definitely be a huge performance gain. A smart one, nonetheless.

Sarmad Burki
Sarmad Burki is a Mathematician and a Economist with a passion for all things gaming and tech. His academics and professional experience combined with tech and gaming adds to his skills giving him a unique ability to observe the tech and gaming industry from various prespectives.