Moore’s Second Law sucks. Everyone knows Moore’s Law. The idea, first propounded in an Electronics Magazine article 50 years ago by Intel Corporation (NASDAQ:INTC) co-founder Gordon Moore, is that semiconductors can get twice as dense, and twice as fast, every 18 months, doubling their price-performance.
Moore’s Law drives the technology economy, but few acknowledge what I call Moore’s Second Law, which is that the cost of the gear needed to make chips will also multiply as circuit densities increase.
Moore’s Second Law has driven the vast majority of chip companies out of the manufacturing end of the business. It needs to drive Intel out, too.
The way to do that would be to split the company into two parts. One would be dedicated to designing and selling semiconductors, which is mainly a function of software and dealing with customers who design products. The other would be dedicated to making them, which is a function of capital and dealing with customers that need chips. It’s the one logical step Intel management has fiercely resisted … that it has flatly rejected.
Shareholders have suffered enough.
Consider ARM Holdings
ARM Holdings plc (ADR) (NASDAQ:ARMH) competes directly with Intel in microprocessors, but doesn’t really compete with it at all.
The British company was acquired by Softbank Group Corp (OTCMKTS:SFTBF) in September for $32 billion, but for many years before that it had been running rings around its larger rival, delivering shareholders a 10-year gain of over 950% while Intel’s performance ware barely half that of the NASDAQ Composite.
The secret of ARM’s success was that it only sold chip designs. Other companies, like Apple Inc. (NASDAQ:AAPL), bought the designs, tweaked them to their specifications, then had companies like Samsung (OTCMKTS:SSNLF) manufacture them. The customers got a custom design for a competitive price.
INTC stock had similar problems in the mobile space, where Broadcom Ltd (NASDAQ:AVGO) ran rings around it for years. Intel tried to sell chips and showed Chinese OEMs software that would work with the chips. Broadcom sold designs the OEMs could quickly get to market and sell. Since Broadcom, like ARM Holdings, wasn’t responsible for the economics of manufacture, it wasn’t selling a one-size-fits-all solution.
As PCs have become less important, INTC has suffered from its lack of flexibility. Profit margins have plummeted and debt levels have increased.
INTC stock has become a yield stock, one bought for its dividend, currently yielding 3.1%, but for the quarter ending in July, that dividend ate an alarming part of Intel’s earnings. A crisis is at hand.
New Markets Require New Thinking From INTC Stock
CEO Brian Krzanich, who took command of the company in 2013, is an Intel lifer and steeped in the corporate culture. He even holds a patent for semiconductor processing.
He announced a new strategy in April that was heavy on buzzwords like cloud, Internet of Things and 5G, but light on details.
Intel stock continues to lead in processing chips used in cloud data centers, but it’s facing new competition from graphics chips, an unexpected direction. Graphic chip industry leader NVIDIA Corporation (NASDAQ:NVDA), which like ARM does not own a fabrication plant, is up 144% just in the last year.
Intel is not the industry leader in communications, either. That honor belongs to Qualcomm, Inc. (NASDAQ:QCOM), whose Internet of Things strategy has been to buy NXP Semiconductors NV (NASDAQ:NXPI) for $47 billion, about half its own valuation. It’s the kind of big bet Intel would never engage in.
The Bottom Line for Intel Stock
Intel’s past is that of a semi-monopolist that controlled its market. Its future is that of a competitor whose designs must win against huge, nimble competitors.
The interests of design are software interests. The interests of fabrication are hardware interests. These interests were once in concert for Intel, back when Microsoft Corporation (NASDAQ:MSFT) dominated the device space. Those interests are now in conflict.
Intel stock’s performance is lagging its competitors because fabrication requires enormous amounts of capital and its design business is focused on feeding that capital beast. But fabrication can survive on its own, because there is far less competition than before. The software side of Intel needs to be free.
Intel needs to be split in two.
Dana Blankenhorn is a financial journalist and author of the science fiction story Into the Cloud. Write him at danablankenhorn@gmail.com or follow him on Twitter at @danablankenhorn. As of this writing he owned shares in INTC, AAPL and MSFT.
