Flash drives in the news--and in real life
Wednesday morning I visited the new Menlo Park, Calif., headquarters of Zonbu, makers of the low-cost, service-supported Linux computer I mentioned last month (here). I met with Zonbu CEO Grégoire Gentil, who gave me an overview of the company's business plan and a demo of the system. He also offered to send me a review unit, so I'll have more to say on this subject later.
As I mentioned in that earlier blog post, the Zonbu system comes with a 4GB flash disk in the form of a simple commercially available Compact Flash card. This is perhaps a little more interesting than it sounds. In the old days, one technique for reducing the cost of a computer system was to boot and run software stored on a server on the LAN. Such systems were called "diskless workstations" (I used one of these for a while at my first engineering job in Silicon Valley) or "network computers."
But that was in the days when a computer needed more local storage for the operating system and key applications than could be provided by non-volatile semiconductor storage. Over the last decade, the size of this basic software load has grown a lot--but the capacity of flash chips has grown faster. So now it's practical to provide a small, affordable flash drive to hold all the basic software a computer needs.
Diskless workstations died out when hard disks grew in capacity and dropped in cost enough to provide a positive return on the investment. Now that flash drives can be justified the same way, I think we've seen the last of the "network computer" concept in its original form.
But I don't think there's any risk that flash drives will replace conventional hard disks. As long as Moore's Law works for read/write heads as well as it works for transistors, hard disks will continue to deliver more capacity for the same amount of money, simply because it's so much less expensive to coat a square inch of aluminum with a magnetic alloy than it is to make a square inch of flash memory.
Michael Kanellos published a story on his CNET blog Wednesday (here) about Seagate's plans to enter the flash drive business. In the story, Seagate CEO Bill Watkins is quoted as saying that flash drives might be able to take over 7 percent of the disk-drive market. I couldn't possibly be that specific, but that sounds like a reasonable estimate.
I suppose Seagate can't bear to give up 7 percent of its own market to competitors from the semiconductor market such as Samsung and SanDisk/Toshiba. But these companies will make their own flash drives, and it's going to be difficult for Seagate to compete against them.
The problem boils down to whether Seagate can add more value to a flash drive through its knowledge of interface technology and storage-management software, or these semiconductor companies can add more value with their fabs. Seagate will have to find an affordable source of flash chips; the guys with the fabs will have to acquire or develop the skills to turn chips into complete storage devices.
I can't predict who'll end up with the stronger position, but I can be fairly sure that the competitors haven't revealed--or possibly even determined--their full strategies.
Peter N. Glaskowsky is a computer architect in Silicon Valley and a technology analyst for the Envisioneering Group. He has designed chip- and board-level products in the defense and computer industries, managed design teams, and served as editor in chief of the industry newsletter "Microprocessor Report." He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure. 
It works fairly well but there are a lot of things to consider, and a lot about the model that is still evolving.
If you want to read about my experiences and discuss the Zonbu with others, check out my blog:
http://mrzonbu.wordpress.com/
-Mr. Zonbu