The rumors were true. In a completely out of the ordinary event called ‘Scary Fast‘, translated into Spanish as ‘A savage’, Apple has updated this Tuesday part of its line of Mac computers, specifically the MacBook Pro and iMacand has also presented the new and promising Apple Silicon M3, M3 Pro and M3 Max processors.
We are facing a third generation chipset that marks an important change in the way the company presents its innovations, at least on this occasion. Previous generations arrived in a staggered manner, but now we find ourselves with the landing of the three proposals simultaneously.
Performance for every need, all in the expected 3 nanometers
It is no secret that processor manufacturers are involved in a race to significantly improve their photolithographic processes. And, although nanometers have long ceased to be a strictly technical measurement and have become an important marketing tool, they are a very valid reference point.
So many of us expected Apple to make the leap to 3 nanometers in its computer chips, mainly after reaching an unprecedented agreement with TSMC in the industry for the foundry to pay for chips that will not work well and seeing the new A17 Pro of the iPhone 15 Pro be produced under this process.
Said and done, the Cupertino company has taken this step, a step that comes accompanied, at least on paper, by a spectacular performance and indirectly as a blow towards the new Snapdragon Qualcomm’s X Elite, which are estimated to be up to 50% faster than the M2 in multi-core performance.
Although, as we say, we have seen this photolithographic process on mobile devices, it is the first time we have seen it on personal computers, which is another advantage over the rest of the competitors. What Apple has done, through TSMC, is use extreme ultraviolet (UVE) photolithography machines.
The most significant advance of the third generation of M chips comes from the graphics processing unit (GPU). This is a series of substantial changes implemented at the level of graphic architecture thanks Dynamic Caching. We will have to familiarize ourselves with this term that moves from what is traditionally used.
Typically, the amount of local graphics memory that is allocated to tasks at compile time is determined by the software. This causes each task to access the same portion of memory, depending on the most demanding task, which affects the overall performance of the GPU. Apple has implemented another approach.
How have you done it? Memory is now distributed dynamically in real time at the hardware level. In theory, this promises to thoroughly utilize only the amount of memory you need the task in question. This would be a very important benefit for running processive applications and even games.
There are also other cool features like the Mesh Shading hardware rendering system. The goal is to improve the chip’s ability and efficiency to process the visual geometry of complex scenes. There is also hardware-accelerated ray tracing, something never seen before on Macs.
As in previous generations, we have performance cores and efficiency cores. Performance cores promise to be up to 30% faster than those present in the M1 chips and 15% faster than those in the M2. At the efficiency level, the jump is 50% and 30% faster, respectively.
The new chip architecture can deliver the same CPU performance in multi-threaded tasks as the M1, but with half the power consumption. As we can see, Apple remains focused on maintaining notable energy performance, something that is mainly appreciated in laptops. Let’s look at the characteristics of each chip.
Apple Silicon M3: 8 CPU cores, 10 GPU cores. Up to 65% faster performance than the M1.
Apple Silicon M3 Pro: 12 CPU cores, 18 GPU cores. Up to 40% faster performance than the M1 Pro.
Apple Silicon M3 Max: 16 CPU cores, 40 GPU cores. Supports up to 128 GB of unified memory. Performance up to 80% faster than the M1 Max.