Analysis For HDD manufacturers the whirlwind, the silicon sandstorm, has literally come out of the desert, the Iranian desert as it happens. In the shape of STEC, a booming Californian company with greatness being thrust upon it, the hard disk drive (HDD) triumvirate of Hitachi GST, Seagate and Western Digital face the biggest disruption they could imagine with, potentially, their products being replaced by flash memory-based solid state disks (SSD).
Other HDD manufacturers such as Samsung and Toshiba have flash memory product capability. The three above do not.
Brothers Manouch and Mike Moshayedi came from Iran and started up a structural engineering consultancy in Santa Ana, Orange County, CA, in 1990. Then they switched and started a business, Simple Technology Inc, designing and selling computer memory modules. Younger Mark Moshayedi, an electrical engineer, joined them in 1993.
Simple Technology prospered and bought Cirrus Logic's flash controller operation in 1994. This marked its entry into the flash memory business and it started selling its first flash drive products in 1995. These were for consumer electronic devices. It bought SiliconTech Inc. in 1998 and obtained a business flash memory customer base and operation.
In 2000 the company went public and renamed itself SimpleTech a year later. Two more acquisitions followed: San Jose-based MemTech SSD Corporation, making SSDs and controller chips in 2005; Gnutek Limited in Great Britain in 2006. The very next year it opened a manufacturing, design and test facility in Malaysia. There are now 250 employees at that site.
In the same year Simple Tech sold its consumer flash business to Fabrik Inc of San Mateo, renamed itself STEC, and focussed on business flash memory products. At that point Mike Moshayedi resigned from the business leaving Mark Moshayedi as president and chief operating officer and Manouch Moshayedi as CEO. They both own STEC stock: Manouch Moshayedi has a 12% stake; Mark Moshayedi has a 4% stake.
The business, headquartered in a pair of smallish offices in a Santa Ana business park, made sales of $190 million last year, with flash SSDs contributing nearly $110 million. Its current valuation is $595.21 million, reflecting its status as the curerent king pin in the flash SSD business.
STEC makes its flash drives from bought-in flash memory chips. It mounts these on printed circuit boards, adds controllers, encloses them and sells them to server and now storage array manufacturers. IBM and Dell take flash drives for servers, EMC takes STEC drives with a Fibre Channel interface for its Symmetrix arrays. Other possible STEC customers are Hitachi Data Systems, HP, and Sun.
STEC is not involved in selling flash drives to notebook manufacturers. It's big win is the provision of controllers with a Fibre Channel interface, meaning its SSDs can fit into a Fibre Channel storage array's 3.5-inch drive slot. No other flash SSD supplier can do this.
Fibre Channel drives reputedly have higher margins than commodity and capacity-centric serial ATA (SATA) drives so any invasion of the Fibre Channel HDD space by flash SSDs could have severe profitability impacts on Seagate and Hitachi Data Systems (HDC) who make Fibre Channel drives.
Western Digital is protected from that by not making FC HDDs. However, a new interface for performance-centric HDDs is SAS (serial-attached SCSI) which is having a faster 6Gbit/s SAS II interface developed, doubling SAS drive interface speed from the current 3Gbit/s. It is generally supposed that the SAS interface will, over time, repkace the FC interface, currently running at 4Gbit/s for HDDs although SAN fabrics connecting FC storage arrays and servers are transitioning to 8Gbit/s. There is no expectation of 8Gbit/s HDD interfaces being developed.
Western Digital has a desktop SAS HDD product line which is expected to be developed into enterprise products. Flash memory supplier SanDisk is expected to add SAS interface SSDs to its current SATA SSDs. We might also expect STEC to add a SAS interface to its controllers and then the stage will be set for Western Digital to face flash SSD competition as well.
There is a race going on between HDD capacity and flash SSD capacity with cost/GB being one parameter, the number of IOs per second (IOPS) being a second, electricity needs a third, and write endurance being the fourth. Flash memory only has a finite number of write cycles before its performance degrades. Hard drives are not so affected. Flash SSD controllers need to compensate for the restricted write cycle life with wear-leveling algorithms to distribute the I/O writes equally, or as equally as possible, across the cells in a flash SSD.
It is expected that, as flash SSDs add more bits to their cells in multi-level cell (MLC) technology that cost/GB will go down and capacity will go up, a 4-level cell holding four times as much information as a single level cell.
Seagate is bringing out its own flash SSD line and has sued STEC for patent infringement; a suit which many commentators view as frivolous and which STEC is vigorously defending, having retained Los Angeles-based law firm Wilmer Cutler Pickering Hale and Dorr LLP to fight the suit. It is also reported that Seagate and STEC have talked about a potential acquisition which STEC dismissed.
Kingston Technology is readying its entry into the flash SSD area.
Kingston, SanDisk and Seagate are large and established companies. Their entry into the enterprise flash SSD market could represent disruption to STEC. If it wants to cash in on its current SSD kingpin status then now is the time. As soon as SAS II flash SSDs arrive from its competitors then its market position will be under attack. In fact it may face possibly the first real competition it has faced in its so far charmed life.
On the other hand this is a company that has played its technology and product cards well, very well indeed. It convinced EMC and that is no easy task. We should by no means expect STEC's sizzle to start shrinking any time soon.
Copyright © 2008, Blocks & Files.com