Dare to predict the future

Since the shutdown of the company I was working at previously, I have had some time to think about what I want to do. I still have the desire to start my own company, but it’s indeed getting harder and harder as I have certain family obligations. Next best thing is to join a startup, but there aren’t too many left standing in networking industry. As I have mentioned before, best case scenario would be to join a startup in web application space, but because I don’t have relevant prior experiences, it is tough. That leaves joining a well-established company in networking industry. While that’s good for supporting the family, it is bruising to my ambition and dream of starting a company of my own. However, life would not be worth living without a dream. We live one life, and I just refuse to accept that I will live the rest of my life as nobody in the sea of big corporations. Well, not exactly. My personality wouldn’t let me be nobody. I am too ambitious to be nobody in any company. Either I make a difference and become somebody or get the hell out. I think I’ve always enjoyed limelight – there is something about being a center of attention.

At any rate, I decided to examine history of technology innovations (as far back as I could remember), and see where the future innovations would happen. As people grow older, they tend to have “locked-in” view. I can feel that I already do that sometimes. But, it’s important to have open mind and views. I will let my imaginations go wild, and see where things might be headed in the future. This would be a good exercise.

I would divide Internet into two large areas: physical and logical (or Layer 1 to 4 and Layer 5 to 7 of OSI model). In physical side, you have telecommunications and data communications equipment that is responsible for delivering bits (zeroes and ones) from point A to point B. In Layer 1, most long hauls, WANs and LANs are either fiber or moving to fiber. To desktop is most likely to remain copper, and to laptop is most likely wireless (Wi-Fi variation). Long hauls use mostly SONET, the good-old, reliable SOB. DWDM stuffs much more information into single strand of fiber using different wavelengths. So, as long as you have fiber in the ground, DWDM or any other future boxes will be interested in stuffing more and more bits into the single strand of fiber. Similarly, Ethernet evolution map is pretty much intuitive, too: faster and faster. Also interesting thing is moving everything over IP. Storage used to be exclusively on fibre channel network, but with NAS and iSCSI, everything is accessible over IP without specialized network. There is even talk about implementing fibre channel over Ethernet. Actually, much more interesting things in physical side are happening in and around data centers. The need to have scalable data center is pushing for HPC environment where resources from multiple servers are pooled together. What’s happening now is like creating a humongous server with hundreds of CPUs and obscene amount of storage. With pre-partitioned storage, any number of CPUs can be instantly grouped together to perform certain jobs. For example, if there is sudden surge of demand for database processing, addition CPUs can be assigned to already existing database CPUs. It is like dynamic server virtualization. You can also imagine, with fast enough connection and fast enough storage like maybe Solid State Drives, that there could be separate “memory area network” in addition to “storage area network”. Thus you have three physically separate areas – CPU, memory and storage – being grouped dynamically on the fly and providing services to clients as needed. What provides physical connections for those three areas become quite interesting too…. Memory typically requires 50ns or less access speed, so I am not sure if current Ethernet switch can work. But, you could imagine some sort of box providing network connection to/from clients as well as between the three areas. That would be really neat… But the box needs to be as scalable as the computing resources. Also management would not be easy. There may be additional challenges I am not seeing right now, but I would believe that’s where most network, server, and storage vendors are heading.

In terms of physical side of mobile industry, it’s also pretty much predictable. Apple iPhone and Google Android are paving a new era of mobile networks. Service providers need to upgrade their equipment to deal with more and more data. Thanks to the two pioneers, mobile phones will be considered as mini-computers where consumers are free (as much as phone manufacturer lets them) to download and install applications.

Logical side is even more interesting than the physical side in both wired and wireless networks. Whether you use desktop or mobile phone, what you do with bits delivered via the network is where the true value resides. However, I must say that I am much more excited about mobile apps than desktop apps. The evolution of desktop apps seems quite predictable. For example, social networking websites could be considered as enhanced BBS. I remember when I first got my computer in high school, it came with 2400 baud modem and the only “online” activity would be through BBS. I exchanged games with others physically…using 5.25” floppy disks. You could choose which BBS to go and hang out. When I went to college, I used newsgroups and IRC as BBS. Then to Yahoo Groups, and now it’s Facebook. Truly remarkable development has been around software-as-a-service model. In most cases, big software vendors tended to target customers who leave most margins, i.e. Fortune 1000 companies. Small-to-Medium businesses usually get crippled version at discounted price, but in terms of productivity, it could be considered a lot more expensive. When you move apps to web, now you have different economies of scale, and distribution and pricing model. Just as the Internet made “Long Tail” possible, SaaS changes the whole software landscape and makes it attractive to SMBs (not that it wouldn’t be also attractive to larger companies, but they may not have compelling reason to jump).

Also, when you add mobility, you get a whole different set of software. Mobility means your location may change at any time. The obvious apps are the ones that tell you about things around you, whether you are looking for a restaurant, a friend, etc. There is also notion of instant social gaming, where you hook up with whoever is available and play a game together. Another one is instant access to information wherever you are. One app I saw lets user scan a barcode of a product and find review/rating information about it. Pretty clever. So what makes apps on mobile phone with high-speed internet connection a lot more interesting is location + instant access to information.

What could be possible? Where could things go from here? SaaS means both desktop and mobile phone could access the same application. So the SaaS should be able to accommodate information from/to both desktop and mobile phone. Will desktops become just another (immobile) terminal to apps? How about SaaS of SaaS? If apps are moving to web, and there might be a need for information exchange between two or more SaaS apps. Mashup for mobile apps. There is an idea! Another characteristic of mobile phone is it’s most likely to be with owner all the time. So, it could be use as tracking device…..could be as physical as distance travelled or expenses……or some sort of analysis based on accumulated data…..like when s/he is most likely to spend money, etc. I actually have envisioned a society without cash, since a lot of financial transactions already happen without me actually touching the money. Money gets deposited using direct deposit. I always use credit cards for purchases. I then pay for credit charges using online banking. I don’t need to touch the money, period. So, what if you add charging capability to the mobile phone? It’s already done in Asia and Europe, where you can pay for goods using your mobile phone. It’s just that mobile carriers are not credit card companies, but it should be possible perhaps through partnership. Then SaaS could keep track of most of expenses through the mobile phones and provide you with financial analysis. That could be possible.

Actually, it would be impossible to think of all the possible mobile SaaS right now….., but it gives good topic ideas. From today on, I am going to write about at least one mobile SaaS a day.

Leave a Reply

Your email address will not be published. Required fields are marked *