In the next 10 years, Intel will have to take the courage to get a chance.
Editor’s note: This article is from the micro-channel public number “DoNews” (ID: ilovedonews) , Author: maomaobear.
Recently, due to a performance that could not satisfy the market, the US chip giant Intel closed down 10.58% on Friday, and its market value evaporated by US$26.56 billion (about 177.5 billion yuan).
On October 22, Intel announced the performance report for the third quarter of fiscal year 2020. The financial report shows that Intel’s revenue for the quarter was US$18.33 billion, a year-on-year decrease of 4%, slightly higher than market expectations in July; net profit was US$4.28 billion, a year-on-year decrease of 28.6%.
IntelChief Financial Officer George Davis said that the demand for personal computers is shifting from desktops and high-end commercial PCs Entry-level consumer and educational products. Despite the increase in sales volume, the average selling price fell, which affected profitability.
Intel, as the world’s most advanced technology company, has been around for decades. In the past few years, Intel has been ridiculed by netizens while squeezing toothpaste while changing interfaces to make a fortune.
In 2020, AMD’s Zen3 will surpass Intel in an all-round way. In the artificial intelligence computing market, nVIDIA is aggressive, and even Apple has abandoned Intel on desktops. What has Intel done wrong over the years?
Why did Intel miss the mobile era
Intel entered the mobile era very early, and smart phones were not actually launched by Apple in 2007. In 1994, IBM produced smart phones. The first large-scale popularization of smart phones was in 2003.
In the first large-scale popularization of smart phones, Intel was an important supplier. While other manufacturers were still using ARM public version processors, Intel was already designing its own micro-architecture.
However, Intel is not optimistic about ARM processors. In 2006, Intel sold the ARM processor, and made its own X86 low-power processor ATOM.
In 2007, Apple released the iPhone, followed by Android in 2008, and the revolution in the mobile era broke out. Apple purchased Samsung processors, and the Android lineup has risen to Qualcomm and MTK.
When Intel came to understand, it acquired Infineon in 2010 and later entered the mobile market. This mobile market was already dominated by ARM processors.
At this time, Intel can’t pull its face down, buy ARM’s authorization, and force the X86 processor, but Android and IOS are not supported, and the special version of Android has poor compatibility.
Intel spent tens of billions to prepare subsidies for share in the tablet market, but ultimately did not insist.
As a result, the mobile market was lost, and orders for hundreds of millions of pieces a year were given to TSMC, Samsung. TSMC and Samsung have the money to rapidly develop their processes and catch up with Intel.
How did Intel miss artificial intelligence
About 2008, Intel was preparing to launch Larrabee, which combines 3D graphics and general-purpose computing. However, progress in product design and driver development was slow, and Intel gave up again and prepared to make Phi computing cards.
This Phi computing card, because there is no graphics card to share the cost, the price is very high, the scope of application is small, and the supporting software has not been developed, and finally gave up.
Just after Intel gave up, digital currency rose, graphics card mining was developed, nVIDIA and ATI made huge profits.
Later, with the rise of artificial intelligence, the artificial intelligence chip required for high-performance computing is still nVIDIA, which has been preparing for many years to occupy the market.
What makes Intel embarrassing is that the original processor is enough for ordinary computing. For real high-performance computing, everyone uses computing cards with the same graphics card to calculate. Intel is also behind in this part of the ecology.
The foundry of graphics and GPU has also given TSMC, Samsung, TSMC and Samsung have a faster angle to catch up with Intel.
The fatal blow, AMD uses TSMC
In the emerging markets, Intel gave up one by one, while in the PC, server, and notebook market, Intel also has new problems.
The mobile market and GPU have raised TSMC. TSMC has reached 16nm, and it has kept pace with 14nm Intel. Today it is 5nm vs. 10nm. Although Intel’s 10nm is a bit better, it is only equivalent to TSMC’s 7nm. There is still no way to fight back against 5nm.
Moreover, Intel’s 14nm is still the mainstream.
As a result, Intel’s old rival AMD used TSMC’s technology, while the architecture continued to improve, and finally achieved a comprehensive surpassing of Intel.
Now, the market where Intel started is also seriously threatened. It is not surprising that the stock price has fallen because of small profits but quick turnover.
Is there any chance for Intel?
For Intel, although the stock price has fallen, the financial figures are still very good. There are many opportunities.
Intel’s top priority is to put aside its face and use TSMC’s 5nm and 3nm OEMs for high-end servers and mobile processors. At the same time, it is stepping up its architecture research and maintaining its performance advantage over AMD in order to maintain the high-priced basic market.
Then, GPU manufacturing should be started immediately, taking advantage of Intel’s mastery of the standard, integrating high-performance 3D graphics in its own CPU or chipset, and opening up the computing power of 3D graphics for the poor to do general-purpose computing. Establish your own high-performance computing ecosystem and cannibalize the nVIDIA market.
In the mobile field, Intel has sold Infineon and basically gave up. But you can make a fuss on ultra-small PCs and platforms, and use the X86 Windows ecosystem to eat away part of the ARM mobile market.
Intel’s past mistakes have wasted 10 years. In the next 10 years, Intel will have to take the courage to get a chance.