Toshiba president: 25% of notebooks will have SSDs by 2011
The solid state disk industry may never be ready for prime time, according to data last week from the president of Toshiba, a supporter of the technology. SSD's rate of growth may never catch up, he projected, to the rest of the storage industry.
While the overall market for solid-state disk drives is expanding at triple-digit annual rates, according to an account of a speech on Monday by Toshiba President Shozo Saito from the Nikkei press agency, analysts today are noting that just as twice nothing is still nothing, thrice not very much is a small percentage indeed.
Thus a 313% annual growth rate for the SSD market, by Toshiba's estimate, will leave only 10% of worldwide notebook computers with SSDs installed by 2010, and 25% by 2011.
Saitosan's point is illustrated quite clearly in this full-page graph from Nikkei. The snail's pace curve at the very, very bottom reflects the growth rate of the SSD industry...and this is from a company that was very eager to join that industry last December, and still hopes to earn a lot from it.
Subsequent reports based on Nikkei's account quoted Saitosan as having projected his company will produce 512 GB SSDs sometime in 2009. However, the Nikkei report indicated the Toshiba president may not have actually stated so directly; that instead, he presented math which may have led reporters to that conclusion.
Last year, you see, Toshiba invested heavily in a new type of NAND flash memory called multi-level cell. Unlike conventional RAM, each cell is capable of storing more than one digital state; and with Toshiba's current production of three bit-per-cell MLC with eight possible states (binary digits would be one bit-per-cell with two states), the company's stated goal is to manufacture 128 GB SSDs by next month. The company may apparently be on track to meet that goal.
Saitosan said that one of the benefits of this technology is the ability to step up one bit-per-cell and conceivably double the NAND's capacity. He projected that Toshiba would be able to perform this miracle at least once per year. From that, aggregators evidently speculated he meant that total drive capacity would double every year; and using 2007 as a benchmark for the 128 GB promise, they concluded 512 GB by 2009.
However, had anyone bothered to read to the end of the article (a practice we at BetaNews highly encourage), they would have also noted Saitosan said this doubling of capacity would come with the bonus of a 50% cost reduction. Thus one could just as easily conclude that he meant to say that, rather than doubling existing SSD capacity, Toshiba could simply build new 256 GB drives with half the memory, but with one bit-per-cell added.
And Saitosan also gave a warning about a technical reason why simply expanding capacity ad infinitum might not be possible: Apparently MLC NAND flash wears out, after roughly about 10,000 rewrites. He gives these products an estimated five-year marketing life cycle for most consumers, and conceivably, MLC-based multimedia drives may not see 10,000 rewrites during that period of time. But drives used as primary virtual memory caches for the operating system could conceivably wear out MLC SSDs far sooner than their preferred lifespan.