BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Moore's Law's Over: Don't Buy Intel, Qualcomm Or Western Digital

Following
This article is more than 7 years old.

Beginning in the late 1960s you could bet that every 18 to 24 months the number of transistors on a semiconductor would double while its cost fell 50%. Now that predictable pattern -- dubbed Moore's Law -- is no more.

Does that mean the companies that used to follow Moore's Law -- such as Intel, Qualcomm, and Western Digital (I have no financial interest in the companies mentioned in this post) -- are fated to a future of mediocrity? Or can they build or buy their way to faster growth?

Before getting into those questions, why do a handful of companies hit $10 billion in revenue and keep growing at over 20% a year while most stall out?

The short answer is that they have what I call Disciplined Growth Strategies -- the title of my latest book. Such strategies are chained together from a combination of five dimensions of growth: customer group, geography, product, capabilities, and culture.

A CEO possessed of the right traits is an essential ingredient for developing and deploying such strategies to sustain rapid growth. A case in point is the late former Intel CEO Andy Grove -- who approached growth challenges with intellectual humility; crafted a vision and culture that attracted and motivated top talent; and placed big bets on new growth opportunities.

Perhaps Grove's biggest bet was his decision to switch Intel from a maker of memory chips -- such as dynamic random-access memory (DRAM) -- into a central processing unit (CPU) provider for the Wintel duopoly that supplied the rapidly growing PC and server industries in the 1980s and 1990s.

That decision created a big profit pool for Intel -- but when companies stopped gearing up to go online and ultimately switched to smartphones and tablets, Intel's growth slowed down.

To be sure, Intel and other chip makers such as Qualcomm and Western Digital did not give up on the idea of building new products that would keep Moore's Law moving ahead.

Sadly, in 2014, with manufacturing costs rising and severe technical challenges growing, the cost of individual transistors stopped falling. That ended, at least temporarily, the ability of semiconductor manufacturers to make new chips that comply with Moore’s Law, according to the New York Times.

In anticipation of reaching the limits of Moore’s Law, researchers have developed a range of new technologies for moving electrons. Among these innovations is the idea of representing data in an electron’s orientation in a magnetic field -- dubbed Magnetoresistive Random Access Memory (MRAM) -- rather than by the charge on a capacitor’s common node – as the industry had done for decades.

Unfortunately, the MRAM market lacks the size or growth potential to accelerate these three chip makers' top-lines.

The one publicly-traded MRAM maker Everspin -- whose stock has lost 2% of its value since its October 2016 IPO -- grew 2016 revenues a mere 3% to $26.7 million as it lost $16.7 million.

Everspin -- which is more optimistic than others -- sees an MRAM market that hit $1 billion in 2015 and will grow at a 19% annual rate to $1.8 billion by 2018.

Lyons, France-based analyst Yole Developpement believes that the MRAM market should grow at an 82.9% annual rate from $40 million in 2015 to $1.5 billion in 2021.

And independent industry analyst Jim Handy believes that the MRAM industry will remain relatively small. As he told me, “For now the applications in which MRAM offers clear advantages over traditional DRAM and Flash are fairly small. The MRAM industry will be in the tens of millions or hundreds of millions of dollars – much less than the $30 billion markets for both DRAM and Flash.”

With MRAM falling short, Intel, Qualcomm, and Western Digital seem to be placing parallel bets on growth from both developing and acquiring new products. Successful acquisitions are uncommon and depend on passing four tests: the industry must have significant profit potential and the combined companies must be better off, generate a positive net present value, and be well-integrated once the deal closes.

Intel

Intel's recent financial performance has been mediocre. Its stock price has increased 25% over the last five years compared with a 66% rise in the S&P 500 during that time. Its revenues have grown at a 4.1% three year annual rate -- slower than the industry's 7.3% -- on the plus side, Intel's net margin of 17.4% beats the industry's 14.3%, according to Morningstar.

Intel is working with Micron Technology on a new MRAM-like chip. Called 3D XPoint -- and based on a technology that Intel does not want to share, though many experts think it's Phase Change Memory (PCM) -- Intel plans to bring out a series of products marketed under the Optane brand.

But their introduction has been delayed. An Intel spokesperson said on September 22, 2016, “Intel is fully committed to bringing Intel Optane technology products to market, later this year and in early 2017. Intel Optane products are healthy and on track to unleash vast system performance potential.”

Intel announced on March 19 that Optane SSDs would be launched sometime in the second half of 2017 so the wait continues.

Meanwhile, on March 14 Intel decided to follow Qualcomm -- six months after it made a related deal -- by acquiring Mobileye, a driverless car digital vision systems supplier, for $15.3 billion. To be sure, Intel thinks the market for driverless vehicles could hit $70 billion in the next 15 years but plenty of regulatory and technical challenges will need to be overcome before that happens. What's more, Intel has had a very tough time integrating past acquisitions -- a case in point is McAfee.

Qualcomm

Qualcomm -- with a history of making cell phone chips -- has posted disappointing recent financial performance. Its stock price has fallen 17% over the last five years compared with a 66% rise in the S&P 500 during that time. Its revenues declined at a 1.8% three year annual rate -- while the industry grew at 7.3% -- on the plus side, Qualcomm's net margin of 20.6% beats the industry's 14.3%, Morningstar says.

Qualcomm has developed MRAM through research projects with partners, rather than seeking to develop proprietary MRAM technology on its own. In 2013, Qualcomm partnered with Belgium's Imec research center to jointly develop MRAM technologies -- with help from chip maker Global Foundries.

Qualcomm also partnered with Daniel C. Ralph, a Cornell University physics professor, to develop Spin Orbit Torque (SOT)-MRAM which he told me faces "significant physical and material challenges" that need to be overcome before it can be delivered to customers.

In October, 2016, Qualcomm announced it was acquiring NXP – the largest chip provider for the automotive market – for $47 billion, in order help Qualcomm diversify its business outside of mobile. Qualcomm president Derek Aberle told Forbes, “It brings a sizable new component to our business. We have a strong position in mobile, but NXP brings unique assets in auto and the Internet of Things.”

Western Digital

Data storage technology maker Western Digital's recent financial performance has been mixed at best. Its stock price has increased 84% over the last five years compared with a 66% rise in the S&P 500 during that time. However its revenues fell at a 5.4% three year annual rate -- worse than the industry's 2.9% decline and Western Digital's net margin of -2.7% compares unfavorably to the industry's 0.6%, notes Morningstar.

In May, 2016, Western Digital paid $16 billion to acquire SanDisk, a NAND-flash storage maker. This acquisition"more than doubled Western Digital’s addressable market.”

MRAM is a low priority for Western Digital, which invested in Everspin in 2014 but has not committed to selling MRAM-based products to its customers. Instead, Western Digital is placing its primary emphasis on SanDisk's 3D-Resistive RAM (ReRAM).

These three companies look to me as though they overpaid to acquire growth even as they try to develop a new generation of chips that appear to be falling short of Moore's Law.

A bet on the S&P 500 index would be a better investment.

Follow me on Twitter or LinkedInCheck out my website or some of my other work here