Jim Keller's mission may be the most interesting challenge of his microprocessor engineer's legendary career--reshaping Intel's x86 processor for a new era after 2020.
Partly because of the good fortune brought by binding with the Microsoft operating system, the Intel processor architecture dominated the PC market; After years of hard work, x86 also replaced the deep-rooted Linux architecture server processor in the data center market. Gaining dominance. Although x86, which failed to occupy a place in the iPhone's modem chip today, missed the smart phone boom, it found itself standing at the starting point of a new era—machine learning is rewriting the rules of the computing market and may make micro- The processor area shuffled again.
Past leaders will often be replaced as technology leaders change with the rise of new leaders; To withstand the new wave of attacks, Intel's management and engineers have to resort to all the skills and luck.
So far, Intel CEO Brian Krzanich has led the transformation of this huge 'corporate cruise ship' to perform well in response to the upcoming wave of change; he excavated from Qualcomm at the end of 2015. The Venkata (Murthy) Renduchintala, sharply reorganized Intel's organizational structure. In the past two years, it has successively acquired companies such as Movidius, Nervana, Mobileye, and Altera to allow Intel to build an arsenal of AI architecture.
Now we need someone to figure out what to do and when to put everything in the right place. This is what Keller Tesla left to join Intel at the end of April last year.
Keller has led the design of the Zen series x86 processor cores during AMD's time and is now making AMD return to profitability. In addition, he also participated in the K8 Athlon design earlier; Keller also worked for two new chip design companies. Broadcom has held several positions and has also designed Apple and iPhone SoCs.
But as someone has said, past performance does not guarantee future success; Intel’s traditional cultural drive is a major force that needs to be confronted, and it is still unclear how artificial intelligence (AI) will reshape the computing domain. Other great resistance is playing a role.
Intel is also facing the problem of slowing down the CMOS process speed. The company’s past growth engines and the entire semiconductor industry are showing signs of being old. Even Intel’s own fabs have not been successful in mass production at the 10-nm process node.
At the same time, the chip market is heading towards an era of hyper-fragmentation. No single system can occupy a significant amount of territory in the PC and smartphone space, so future companies have to struggle at the same time in different areas— ─ From Internet of Things endpoints, smart watches to self-driving cars, data center networks, etc. – to strive for design orders.
This is a series of intertwined challenges. There is no doubt that a veteran industry veteran like Jim Keller should have a thorough understanding of this situation; I was fortunate enough to be able to sit in the stands and watch excitedly for the next 3 to 5 years. Before returning home - or turning to the next career challenge - there will be something happening.
Compilation: Judith Cheng