Analysis Despite all the pomp and circumstance surrounding Apple’s move to local silicon for Mac, the tech giant has admitted that the new M2 chip isn’t quite the slam dunk its predecessor was compared to the last of Apple’s former processor supplier, Intel.
During its WWDC 2022 keynote on Monday, Apple focused its high-level sales pitch for the M2 on claims that the chip is much more power efficient than Intel’s latest laptop processors. But in doing so, the iPhone maker admitted that Intel beat it, at least for now, when it comes to processor performance.
Apple made this clear during the presentation when Johny Srouji, Apple’s senior vice president of hardware technologies, said that the M2’s eight-core processor will deliver 87% of the peak performance of the 12-core Core i7-1260P. from Intel while using only a quarter of the power of the rival chip.
A concession from Apple on CPU performance, but the M2 is much more energy efficient. Click to enlarge.
In other words, Intel’s Core i7-1260P is nearly 15% faster than Apple’s M2, and that’s not even taking into account that Intel has two more powerful i7s in its so-called lineup. P-series: the higher-frequency i7-1270P, which has the same number of cores, and the 14-core i7-1280P.
The company claimed that the M2’s processor is 1.9 times faster than Intel’s 10-core Core i7-1255U while using the same amount of power, but while that might be a more apt comparison, the fact is that Apple does not have a processor. for ultra-thin laptops as powerful as Intel’s best.
Either way, Apple says performance per watt, where the M2 really shines, is the most important metric, building on the original argument it made when it launched the M1 in 2020.
“Unlike others in the industry who dramatically increase power to gain performance, our approach is different. We continue to focus relentlessly on energy efficient performance. In other words, maximizing performance while minimizing power consumption,” Srouji said.
But performance-per-watt isn’t the only way Apple hopes the M2 will stand out when it lands in the 13″ MacBook Air and MacBook Pro next month.
The tech giant is also making a bigger bet on the chip’s GPU and neural engine as it believes a growing share of applications in the future will rely on graphics and AI, according to the veteran analyst. Semiconductors Kevin Krewell of Tirias Research.
The M2 is still an impressive chip overall, especially its GPU and neural engine. Click to enlarge.
This is reflected in Apple’s decision to dedicate more transistors to the M2’s 10-core GPU and 16-core neural engine compared to the M1, Krewell said.. These design decisions have allowed the Mac maker to claim a 35% increase in GPU and 40% increase in Neural Engine over the M1. On the other hand, the M2’s processor only improved multi-threaded performance by 18%, according to Apple.
But even then, Krewell said, apps that are highly CPU-dependent, like web browsers, don’t need faster chips as much, which is why he thinks it’s important to tune more. weight to the GPU and Neural Engine since they might make a bigger difference.
“Web browsers don’t need much more performance, so the industry performance comparison is probably less relevant in my mind, although overall power efficiency is good. Apple wants to show that they are competitive with Intel and that in some way they can be ahead with neural processing and better graphics,” Krewell said. The register.
The M2’s GPU looks pretty good – if you believe Apple’s claims. Click to enlarge.
While Apple didn’t provide a competitive comparison for the M2’s neural engine, it did claim that the 10-core GPU is 2.3x faster than Intel’s Core i7-1255U integrated graphics while still using the same power. Conversely, Apple said the M2’s GPU can deliver the peak performance of the i7-1255U while only using a fifth of the power. The caveat is that Apple didn’t provide a comparison to the i7-1260P, which has a faster integrated GPU than the i7-1255U.
By its own admission, Apple may not have the fastest processor in the industry for an ultra-light laptop. But its greater emphasis on the GPU and neural engine lends itself to the growing trend in the computing world that having a faster central brain may be less important than having dedicated accelerators for increasingly important areas like AI and graphics. ®