Look Beyond The Storm Clouds Hovering Over Silicon Motion
Jul. 24, 2017 7:14 PM ET
About: Silicon Motion Technology Corporation (SIMO), Includes: AAPL, HXSCL, SSNLF, TOSYY
Growth at reasonable price, long-term horizon, value
Silicon Motion's moat in the industry is its expertise in developing controllers for NAND flash makers.
The industry is actively migrating from production of 2D NAND flash to 3D NAND flash. The process is complicated and costly.
While the demand for 3D NAND flash rapidly grows, supply is still deficient. These pressures have impacted Silicon Motion in recent quarters.
But this trough will eventually be but a blip in the industry's history.
Silicon Motion ( SIMO) is the leading NAND flash controller supplier worldwide. And, therein lies a problem, albeit a short-term one. NAND flash makers are in the midst of a migration from 2D (planar or single layer of memory cells) to 3D (stacked vertical layers of memory cells) production. Changing the manufacturing process is complicated. The manufacturing process for a string of 2D NAND cells runs longitudinally. For 3D NAND, the string of cells is turned on its end to become a vertical string. Furthermore, the new manufacturing process requires significant investment.
“It’s tough to etch a hole sufficiently deep and narrow, and sufficiently consistent from top to bottom. It’s also very tough to keep the width of the hole consistent while going through layers of very different materials.”
“It’s safe to assume that migration to 3D will require an even larger complement of new materials tools at a significant increase in cost.”
This migration is pressuring the supply of 3D NAND flash. Yet, more and more, the industry is demanding 3D NAND flash to take advantage of the increases in memory capacity.
The tightness in the NAND flash supply did not come as a surprise. The industry was well aware the migration would be slow. Silicon Motion's management, as well as others, started sounding warnings in mid-2016.
"Our growth in 2017, however, will be constrained by limited flash availability with channel SSD. When we and our customer have higher certainty about the roll-out of the new supply, we may look to revise up our guidance."
It was expected NAND flash suppliers would take the first half of 2017 to fine tune their production yield. On April 18, Digitimes projected NAND flash supply pressures would ease in the latter half of the year.
“Samsung, which has relatively stable yield rates for 3D NAND technology, is expected to scale up its output of 3D NAND chips significantly between May and June. Meanwhile, yield rates for other NAND flash vendors' 3D NAND technology are set to improve substantially boosting their chip shipments in the second half of 2017.”
Silicon Motion's moat in the industry is its expertise in developing controllers for NAND flash makers.
It partners with flash makers to develop controller solutions that are optimized for performance, lower power, lower cost and ease of use. With the company strategically ingrained in flash makers' progress, its projections aligned.
“We continue to believe NAND flash supply will increase more meaningfully in the second half of this year and our business will rebound accordingly.”
So, it was hardly a surprise when Silicon Motion's preliminary second quarter results were “off.” With its first quarter reporting, the company projected revenue would fall in a range of $134 million to $140 million. The non-GAAP gross margin was expected to be 48.5% to 50.5%. On July 7, 2017, Silicon Motion reset expectations. Revenue is now expected to be slightly below the low end of the range. Gross margin is expected to hit the low end of its range.
Silicon Motion is notoriously conservative with its guidance. But the company's actual results are consistently inline with preliminary results. The market was disappointed at the warning that revenue would fall below the company's guidance, especially knowing Silicon Motion tends to be conservative. The share price fell over 6% on July 7 and continued to decline.
It didn't matter the company would be returning to quarter-over-quarter revenue growth. In the 2016 fourth quarter and 2017 first quarter, sales had slipped due to the insufficiency of NAND flash supply. Rather, sales less than $134 million in the 2017 second quarter surely means Silicon Motion will experience its first quarter with a year-over-year sales decline since the company's transition from a supplier of external storage in 2013 to the leading supplier of embedded memory and SSD controllers in 2015.
Source: Author-created from company data
Loss in the share price also could have been related to news about Apple ( AAPL) rippling through the industry. On Thursday, July 6, Digitimes publisheda warning concerning a shortage in 3D NAND flash supply for Apple's upcoming release of the latest iPhone. Apple's launch announcement for new 2017 iPhones is expected in September.
“Apple began using non-volatile 3D NAND chips in its mobile devices last year because of the technology's ability to pack more storage space into equivalent dimensions compared to previous flash memory. However production of the chips is a more delicate process, and it looks as if Apple's priority suppliers haven't been able to increase their yield sufficiently to provide for the devices the company is planning to launch in 2017.” (emphasis added)
Based on the notion its priority suppliers, SK Hynix ( OTC:HXSCL) and Toshiba ( OTCPK:TOSYY), would fall short, Apple turned to Samsung ( OTC:SSNLF) to make up the shortfall.
The possibility of the 2017 first half shortfall lingering longer into the second half of the year may well be the culprit for dark clouds lingering over Silicon Motion.
Disparaging looks should not be cast toward SK Hynix or Toshiba. The deficiency in the 3D NAND flash supply is hardly the result of a lack of effort on their account.
In late December 2016, SK Hynix announced it would build a $1.8 billion factory in South Korea to produce 3D NAND. Construction is scheduled to begin in August, 2017. Completion is expected in 2019.
In April 2016, SK Hynix introduced 3D NAND based on a 36-Layer 128Gb chip. Six months later, it was mass producing 48-Layer 256Gb chips. By April of 2017, it had developed a 72-Layer 256Gb chip. Mass production of the latest chip was scheduled for the second half of 2017.
In November, 2016, Toshiba announced it would build a facility in Japan “dedicated to 3D Flash memory processes.” The build was set to occur in two phases with the first to be completed in the summer of 2018. In late June, Toshiba shared it would invest 180 billion yen for manufacturing equipment installation for the first phase. Additionally, construction of the second phase will begin in September and complete by year-end 2018.
On January 27, 2017, Toshiba announced its intention to spin off its NAND flash production and offer a third party no more than a 20% stake. Toshiba Memory Corporation was split off from Toshiba Corporation on April 1, 2017. On June 21, Toshiba shared its board had selected a consortium of Innovation Network Corporation of Japan, Bain Capital Private Equity LP and Development Bank of Japan as the third-party investors.
However, Western Digital's ( WD) acquisition of SanDisk in 2016 presented bumps in the road for Toshiba. SanDisk and Toshiba had a joint venture at the time of Western Digital's acquisition as well as at the time of Toshiba's spin-off. Western Digital claims the transfer to Toshiba Memory should not have been made without SanDisk's agreement. Toshiba claims Western Digital interfered with the bid process for the minority stake in Toshiba Memory. As of July 18, the two companies are attempting to form an arbitration panel.
Despite the structural and legal wranglings, development has continued at Toshiba Memory. In late June, it announced the first 3D NAND flash with a quadruple-level (QLC) stacked cell structure. The company started shipping prototypes in early June for evaluation. The product will be introduced at the 2017 Flash Memory Summit in August.
It could be pertinent to understand yet another factor impacting the industry at this point. Mobile devices incorporate either eMMC, UFS or PCI-E/NVMe flash controllers to utilize flash memory.
“A flash controller provides the interface between the host controller or processor and the flash memory device. It decides where in the flash memory to store the information. It then retrieves the information when requested. Flash controllers will also organise the way data is stored to achieve ‘wear levelling.'"
Silicon Motion supplies all three types of controllers to the industry.
According to an initial technical review in September 2015, Apple's iPhones have incorporated PCI-E/NVMe storage since the 6s.
“The iPhone 6s in turn appears to use NVMe, which rules out both UFS and traditional eMMC.”
At the Flash Memory Summit in August, 2015, Sandisk predicted the eventual adoption of such a solution.
In 2015, Apple's iPhone, as a single product, was the largest consumer of NAND flash at 15%. In 2016, with the introduction of the iPhone 7 and multiple storage options, the product's consumption was expected to increase significantly.
However, Android smartphones, so far, have proceeded another direction. In March 2017, Silicon Motion announced its UFS 2.1 controller family. The solutions deliver “ultra-high random read/write performance of up to 50,000/40,000 IOPS, capacity up to 512GB and ultra-low power consumption.” Silicon Motion was working with a top five OEM (original equipment manufacturer) of Android smartphones to utilize this generation of UFS controllers. It expected the OEM to ramp to production volume by the second half of 2017.
On July 20, 2017, Digitimes released research projections based on the choice between UFS (Universal Flash Storage) and eMMC (embedded Multi-Media card).
“UFS2.1-based NAND flash chips boast the highest data transmission speed of up to 11.6Gbps (1200MB/s) compared to 3.2Gbps (400MB/s) attainable by eMMC5.1 chips. The difference will make UFS2.1 NAND flash chips the mainstream storage technology for smartphones in 2017.”
When 2017 commenced, Silicon Motion expected otherwise. Third-party research had placed UFS penetration at 10% - growing to 25% in 2018 and 45% by 2020.
Whether it is the manufacturing process, the increasing capacity or the controller protocols, there have definitely been developments in the industry surrounding 3D NAND flash in the past year. Supply may still be inadequate but demand continues to increase. It is not at all likely demand would decline to match supply.
Beyond mobile devices, flash memory also is being used in solid-state storage devices. SSD devices are actively replacing hard disk drive technology. Additionally, SSD devices are built for both enterprise applications as well as consumer products such as cameras and USB sticks.
In 2015, Samsung projected SSD consumption of 3D NAND flash would bypass the consumption in mobile devices
Even with the dramatic growth projected for NAND flash, it was projected in 2014 by the IDC (International Data Corporation) and Western Digital that the category would be supplying only 10% of the storage needed in 2020.
So, it is true. There are storm clouds hovering over Silicon Motion because it is reliant on the production capabilities of NAND flash suppliers. Because of the migration pressures in meeting the demand for 3D NAND flash, the company's sales growth has slowed in the most recent three quarters. But development has certainly not slowed at Silicon Motion. The company is still tasked with moving in lockstep with the NAND flash suppliers by having controllers ready as their products improve. And NAND flash makers have plenty of motivation to stay on target.
This trough in the industry's history may feel long lived but it's destined to be but a blip. Silicon Motion's prospects are hardly dwindling or depressed. When second quarter results are announced July 31, Silicon Motion's long-term potential may be overshadowed by short-term conditions. Alert investors will see beyond the clouds darkening the landscape.
Disclosure: I am/we are long SIMO, AAPL.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: I belong to an investment club that owns shares in SIMO and AAPL.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.
About this article:Expand