Tuesday, November 29, 2005

The rise of flash memory 2 comments



(P.S: Sorry for any disturbances the advertisements above may have caused you)
Five months ago in one of my first blog articles (has it been five months already?!) I wrote about the Future of Mobile Phones and discussed the tussle for ascendancy between flash and hard disk drives to become the storage form of choice on mobile phones, or indeed on most personal electronic devices, currently the hottest sector in technology.

It looked more likely then that HDDs might just gain the upper hand, given their success on the Apple Ipod, and observing the miniaturisation trend that was giving birth to 1" and 0.85" HDDs. However, several events have occurred since then to swing the trend, probably decisively, towards flash memory. The chief trigger was the adoption of Samsung's NAND flash memory by Apple in its latest version of the Ipod, the Nano. Apple also more or less declared that flash was its choice for the future. The second event is rather, a combination of events involving the main vendors of both storage platforms. Samsung announced that it was committed to growing the market for flash, revealing its 16GB prototype and plans to expand to 100GB in a few years. Given its financial strength, top ranking in global flash memory compared to only being 4th or 5th in HDD, and its strong and growing downstream Samsung electronics brand, one cannot doubt its commitment nor its ability to bring fruition to that promise. Over on the HDD side, Seagate, the world's top HDD supplier, announced it would cut back on capacity expansion plans, revealing attentuated optimism; Maxtor, probably the 2nd or 3rd largest, announced it was ditching its 1-inch drive development altogether.

As usual, the economics of the options decides the contest. Flash is still much more expensive per unit megabyte compared to HDDs but its price is dropping rapidly; it had fallen about 40% in 2004-5; in general they are declining in prices faster than for HDDs (~20% annually). It will not be surprising, as well, to see the flash vendors, led by Samsung, make a concerted push, at the cost of short-term margins, to drive mass adoption of flash in personal consumer electronics. There are three key personal electronics items: the digital camera, the music player, and the holy grail of the mobile phone. The first two have more or less embraced flash; it is likely that the key vendors of the last will do so as well.

The key advantages of flash are that it is more robust (since it has no moving parts) and consumes less power; these are key advantages in personal devices which are often operating on the go, and the battery unit forms a significant portion of the cost as well as its weight. For mobile phones, the adoption of flash as the standard might be less straightforward: a key problem is that flash has a limited write/erase cycle; on mobile phones where such operations will take place many times daily (projecting into the future when a phone becomes a personal data-organisation device) this will be a serious technical drawback. But I wouldn't bet on this problem remaining unresolved; too often the case is that when the creative human encounters a difficult technological problem, it is the reputation of the former that remains intact.

Don't sound the death knell for hard disk drives. In terms of cost per megabyte HDDs will take some beating, and on static devices such as desktops they will remain the de facto choice. It may be interesting to note that flash vendors are now planning to encroach on the laptop domain; if they do succeed it would be time to dump all your HDD-related shares.

References:
(1) IDG news article May 23: Samsung develops flash-based 'disk' for PCs
(2) Yahoo news article Nov 16: Maxtor Inches Away From Small Drives

 

 

2 Comments:

Anonymous Anonymous said...

I think HDD will extinct within 5 years after 100GB Flash is successfully commercialized.

12/02/2005 6:45 AM  
Blogger DanielXX said...

Indeed, my exact sentiments. But the HDD players are not sitting still -- if their R&D technologies can be successfully commercialised, such as perpendicular recording, heat-assisted recording etc, they could multiply disk capacity several hundred-fold while reducing their size (especially important for consumer electronics).

But a risky bet.... so rather hesitant about HDD stocks, especially if Seagate is going to pass its R&D costs down to its suppliers. It probably will, given its increasingly dominant position.

4/14/2006 10:08 PM  

Post a Comment

<< Home