Basically, we are in a world where “as long as we can have a little connection with Apple, it is a guarantee of box office and article views.” In April 2008, Apple spent US$278 million to acquire PA Semi, a fabless chip design company founded in 2003. Then in March 2010, it spent US$121 million to buy Intrinsity, which specializes in circuit design optimization. After ten years of hard work, On November 17, 2020, the M1 processor came to fruition, giving Macs the battery life and “ready availability” of Wintel laptops.
Apple, which has been “godized” by many Apple fans for a long time, always seems to be “omnipresent and omnipotent”. Mobile phones and personal computers are far from enough. Even the “M1 is coming fiercely, Apple will also enter the server in the future” is a bit overly enthusiastic. Irresponsible predictions do not seem unreasonable.
Advertisement (Please continue reading this article)
▲ Before PA Semi was acquired by Apple, it did have plans to develop a server platform. However, how much motivation Apple has to develop its own server chips is a big question mark.
But looking back, Apple has a very powerful R&D team that can “work hard to create a unique processor core”, but its cloud services are relatively “two-pronged”. If it develops its own data center server chips, is this a cost-effective business? Apart from being a question mark in itself, for Apple, the 5G modem is still worthy of concern. Apple has been trying to develop 5G modem chips, but later found that the massive tactic of mobilizing thousands of engineers could not produce modem chips that met expectations. For the iPhone 17, it still had to rely on Qualcomm (Qualcomm), and there were even rumors that Apple had decided Rumors of abandonment of development.
Advertisement (Please continue reading this article)
Over the years, Apple has been the “pioneer” in TSMC’s most advanced manufacturing processes, spending a huge amount of manpower and material resources. When it broke through the 3 nanometer barrier, it in turn affected the research and development of 5G modems. As for the overwhelming “Apple 3 nanometer chip performance” It seems to be unsatisfactory, and the modular commonality of the circuit layout of the M3 family is obviously not as good as that of M1 and M2.” There is no need to waste space in elaborating.
The difficulties Apple has encountered and the price it has paid in promoting Apple Silicon may far exceed outsiders’ speculation. Whether Apple can still develop server processors is a big question mark no matter how you look at it.
▲ Judging from the chip layout, the “modular commonality” of Apple’s M3 family does not seem to be as obvious as that of M1 and M2. Perhaps the “naturally dangerous” threshold of 3 nanometers leaves Apple’s chip R&D team no room for “elegant stacking of blocks” .
Of course, the trendy “artificial intelligence chip” is also a potential possibility, but the prerequisite is that “Apple must first have a clear set of artificial intelligence strategies.” But unfortunately, we haven’t seen a shadow of it yet. Waiting to see whether Apple will “tell the world” a clearer direction at the evangelistic conference in 2024.
(Source of first picture:apple）
