Apple is suspected of false propaganda, claiming that the 4nm A16 processor is actually a 5nm process

News on August 22, according to foreign media Wccftech reports, although the A16 Bionic chip on Apple's iPhone 14 Pro series claims to use TSMC's 4nm process, the latest news shows that this chip is actually a 5nm chip. Apple's marketing team has false propaganda to mislead consumers.

The well-known whistleblower @URedditor said that Apple's A16 Bionic claims to be a 4nm process, but "I see it marked as a 5nm chip internally." This also explains why Apple's press release mentioned that the M2 Pro and M2 Max processors on the updated MacBook Pro use 5nm process, not 4nm process. However, @URedditor said that Apple's A17 Bionic processor will be TSMC's 3nm process, which will bring greater performance improvements. Previous benchmark data showed that the A17 Bionic's single-core and multi-core performance was 31% faster than its predecessor.

In addition, under the post initiated by @URedditor, some netizens also pointed out that Qualcomm Snapdragon 8 Gen 2 is also based on the "5nm" process, not the advertised 4nm process, although in some cases the chipset shows better performance than Apple's A16 Bionic. Some netizens also claimed that TSMC's first-generation 3nm process may also be just an optimized version of the 5nm process, and N3E technology is actually the real 3nm process.

It is worth noting that in August last year, TechInsights released a report saying that this "misrepresentation" process does exist in the semiconductor market today, and two leading foundries have allowed customers to claim that they have adopted the 4nm (N4) process, but the actual 5nm (N5) process is still used.

TechInsights pointed out in the report that although the MediaTek Dimensity 9000 claims to be based on TSMC's N4 process, TechInsights found that its key process size is exactly the same as TSMC's early N5 process products. In addition, Samsung's own Exynos 2200 processor claims to be manufactured on a 4nm process, but Qualcomm's Snapdragon 8 Gen 1 (which appears in some S22 models) is manufactured using 5nm technology. To avoid customers' desire for one processor (compared to another) to be ahead of the process, the two companies decided to launch the 8 Gen 1 as a 4nm chip. Because Samsung makes both products, it has created a new process called 4LPX, like the N4, but only serving one product. After analyzing the Snapdragon chip, TechInsights found that Samsung's 4LPX process is not physically different from its 5LPE process.

Why do foundries and their customers misreport their transistor progress? In this regard, TechInsights analysis said that the difference between the fake 4nm process and the "real" 4nm process is only 5% optical shrinkage (10% area reduction). Even this small advance broke Samsung's production model, and TSMC took two years to complete. Given the long interval between nodes, foundries may increase speed, power, or yield (and ultimately achieve the same effect) while waiting for the next node. For example, while Samsung's 4LPX has the same dimensions as the 5LPE process, the former may offer some additional advantages to justify its dedication to the new name. But in the absence of any increase in transistor density, the name should be 5LPX to indicate that the process still falls under the category of the 5nm node. But most people use the abbreviation of 5nm or 4nm instead of the full process name, so the numbers are still important.

Editor: Xinzhixun - Ronin Sword