[ad_1]
By the time Apple removed the wraps from its latest Macs this week, the highlights of the new machines were way ahead. They run on chips made by Apple itself, a change after 15 years of relying on Intel.
However, where Apple surprised, and even surprised, was in the claims it made for its custom-designed silicon. A MacBook Pro with the new M1 chip, for example, is said to be able to handle 20 hours of video playback on a single charge, twice as long as an equivalent Intel-powered machine.
Those claims have yet to be independently tested, and the move to a new chip architecture is no small undertaking given the need for an entirely new software library.
If Apple is proven correct, it will have gained an unexpected advantage. You will also have turned your attention to the new battlefield that has opened up among major technology companies as they seek custom chips to handle their most important applications.
Since the PC era, which saw the emergence of specialty chip companies, hardware manufacturers, and software companies, computing has been largely defined by a horizontal structure. Along with the expertise required to design increasingly complex processors, chipmakers faced the high capital costs of keeping up with the latest manufacturing technology.
Several things are forcing this deeper level of vertical integration into the world of technology. One is that computing needs are changing. The last decade has been defined by the rise of smartphones and centralized cloud data centers.
This is ushering in the era of big data, artificial intelligence, and the Internet of Things. Computing demands are skyrocketing as tech companies look to put AI in every device. Boosting performance while reducing power consumption, always a key issue for chipmakers, has become critical.
In this world, custom chips are the fastest way to gain a competitive advantage. Tesla, for example, is betting that it can produce a fully autonomous car using nothing but AI to analyze images from the car’s cameras. Most of his rivals think this is too ambitious. Yet with the equivalent of a supercomputer equipped with internally designed chips in the trunk of each of its cars, Tesla believes it has a chance to prove them wrong.
Nowhere has this been more evident than in the data centers of the giant Internet companies. Amazon signaled its intention five years ago with the purchase of chip startup Annapurna and now manufactures processors for use in cloud computing centers. Google is already in the fifth generation of its ultra-fast artificial intelligence chips, while Facebook has been working on something similar.
China’s major tech companies, fearing loss of access to American chip technology, have an additional incentive to enter the silicon business. Alibaba, for example, showed off its first AI chip last year, along with one designed to be used in the new equipment required by the Internet of Things. The latter is based on the Risc-V open source architecture. Choosing technology that is freely available like this, rather than the most widely used designs that must be licensed by ARM, is a sign of how Chinese companies are trying to protect themselves.
Another factor that has contributed to this new verticalization of technology has been a change in the structure of the semiconductor industry itself. The emergence of specialized chipmakers, led by the Taiwan Semiconductor Manufacturing Company, has made it easier for newcomers to produce their own silicon. Apple’s new M1 chips, made by TSMC, are based on a manufacturing process that is more advanced than anything Intel can match.
But producing a chip involves much more than design and manufacture. In the middle is what Matt Murphy, CEO of Marvell Technology, calls the “meticulous, nuts and bolts” work of turning a piece of silicon fresh from the manufacturer into a final product, including packaging, assembly and tests. That has led companies like Marvell to seek a new business working alongside the big tech companies.
Google may do much of the work designing its AI chips, but it relies on Broadcom to provide all of this additional chip-making know-how. In the same way, the SQ2 chips that Microsoft makes for a model of its Surface tablet are “co-designed” with Qualcomm.
It’s unclear how far companies like these will go to compete head-on with established chipmakers. But judging by Apple’s breakthrough this week, a deeper understanding of silicon has become a major differentiator.
