Analysis

posted on 04 July 2008 12:48

Worries expressed in a note by an Avian Securities financial analyst about STEC's competitive situation were dismissed by STEC's VP for business development, Patrick Wilkison. Text from the note is reproduced below with inserted comments.

"The stock has appreciated significantly over the last few months largely on the back of an unannounced, but well telegraphed design win for STEC's MLC based SSD in the Apple Mac Book Air."

Wilkison said that STEC policy was not to mention its customers until its customers had mentioned STEC publicly as a flash SSD supply source: "Apple has not said that, so we stay mum."

The Avian note said: "While we expect further good news in the near-term in the form of an additional partner announcement, we are increasingly concerned about STEC's competitive landscape beyond the next six months."

While not revealing specific customers Wilkison offered this thought: "On the enterprise side we have every major OEM in qualification. EMC is public."

The Avian note continued: "STEC's exclusivity agreement with EMC comes to an end in Q3, allowing STEC to pursue other enterprise accounts but also allowing EMC to choose additional SSD suppliers. IBM, Hitachi, and Sun are all slated to introduce SSDs in the 2H08; we believe STEC will likely name one of these as a partner over the next few months. This will, however, be the last bit of good news for the stock."

Wilkison believes that: "There is no alternative source in the enterprise SSD market for the Zeus SSD." He's likening the enterprise SSD market to the enterprise hard disk drive (HDD) market and suggesting that although second sourcing or multiple sourcing of HDDS is common by storage array OEMs it is simply not practical, for now at least, in the SSD area because there is no equivalent to STEC's Zeus enterprise-class SSD available from any other supplier.

The Avian analyst wrote: "We expect INTC (intel) , MU (Micron), SNDK (SanDisk), and Samsung to launch expanded SSD offerings creating an inherent price disadvantage for STEC which lacks internal NAND supply (STEC has yet to partner with a NAND supplier.) Currently all of STEC's NAND is obtained from Samsung."

Wilkison has two responses here. NAND flash foundries and SSD suppliers owning foundries, or with investments in them, are focussed on volume and that means the retail flash market, not the enterprise SSD market.

Wilkison points out: "We don't feel we need to own a foundry. ... The market is still nascent. ... We don't want to make an investment into a fab (as) we source media through very tight collaboration and partnerships. ... If you do invest in a fab then you focus on high-volume and the only volume is in retail ... and that becomes the core focus of the business ... SD cards, etc.... We don't want to dilute our focus."

Avian's note said: "On the enterprise SSD side several smaller competitors such as MTRON should gain traction with solutions that are almost as fast, arguably more elegant, and expectedly cheaper than those of STEC further pressuring the competitive landscape. "

Wilkison's response is this: "The research is a bit shallow (and it) doesn't understand what it takes to support enterprise SSD. It is like the hard drive industry with Seagate, Fujitsu and Hitachi the only suppliers supporting enterprises. Western Digital and Toshiba don't offer enterprise drives. Enterprises have unique needs; it's not one SSD for all applications."

In other words, MTRON and other SSD suppliers will not find a ready entry into the enterprise SSD market. Wilkison thinks that, for flash NAND foundry owners and investors: "SSD is a relief valve" or rather, it appears to be a relief valve though which they can pump their NAND inventory to keep the fabs operating but that appearance is not reality because enterprise SSD needs are different, substantially different.

Wilikison adds: "The bulk of NAND SSD will be in notebooks. It's difficult to substantiate (the) investment to support a vibrant enterprise SSD business (and this is) a barrier to entry."

Intel and Samsung are tuning their flash for notebooks and that takes precedence over server and storage array flash.

MTRON has recently announced a faster flash SSD controller, one with eight channels. Wilkison comments: "STEC controllers already have eight channels and more with volume shipments taking place now." He characterizes the MTRON announcement as defensive as the MTRON product won't be shipping until 2009. Also: "Marvell is bringing out an 8-channel controller out soon, so putting MTRON under pressure."

STEC is taking the view that it is the only practical and viable enterprise storage array flash SSD supplier. Other would-be suppliers will find their product strategies skewed towards consumer and notebook flash because of foundry needs for volume. They will also find that enterprise flash SSD needs are far more comprehensive and demanding than consumer and notebook flash, providing an unexpected barrier to entering the enterprise flash SSD market.

No doubt some of them will overcome it but, as in the HDD market, multiple sourcing will be the rule and, in a growing market, there is room for more than one supplier.

The Avian analyst, in STEC's view, did not understand these things and we should not expect that rules that apply in the consumer and notebook flash market will apply in the enterprise. STEC is not facing a long-term competitive disadvantage at all. It is the current enterprise flash SSD market leader and has, Wilkison implies, the technology expertise and focus to continue developing its products and hold on to that position.

^{[Chris Mellor.]}