July 30, 2007

The ‘Grids in Financial Services’ Two-By-Four

Derrick Harris

Have we made it clear over the past year that grid computing is huge in the financial services market? Have we exposed you, our loyal readers, to enough quotes from financial end-users and vendors touting just how important they are to one another? Are you punch drunk from absorbing too many blows along this front? I certainly hope not because, if so, this week might put you down for the count, as we whack you in the head once more with our “grid in financial services” two-by-four.

However, rather than hear the too-sunny take from a vendor or the carefully crafted company line from a user, this week we have a column from someone who bridges the gap between the two. Marc Jacobs, who spends his days with Lab49 building “advanced applications for the financial services industry,” shares with us his insights into how distributed computing is affecting the financial services industry — and it shouldn’t be surprising to find out that the technology is having a big impact. Taking a look at the past, present and future of grid computing technologies in the market, Jacobs illustrates how far we’ve come and how far we’ve yet to go. While he makes it rather clear that technology to develop high-performance distributed apps currently exists, there are still hurdles to clear. Anyhow, I don’t want to give the whole thing away here, so please take a look for yourself.

This week’s issue also includes an interesting announcement from United Devices that shows just how the company is looking to redefine “grid computing” with its line of solutions. Rather than simply connecting PCs, the new UD is all about optimizing the datacenter and bringing to it new opportunities for innovation, and its work Japan’s NTT West is a prime example of this. Essentially, by deploying UD’s managed services offering, the telecommunications provider is able to offer its broadband customers the ability to sell CPU cycles into NTT’s Hikari Grid, which, in turn, provides computing resources to the company’s business customers. There has been plenty of talk over the past few years about how badly telcos want to offer grid services to customers, and in this case the telco doesn’t even have to maintain the entirety of the hardware resources it makes available. It’s like volunteer computing meets utility computing, and I, for one, love the idea.

Elsewhere on the Web, in the blogosphere, to be exact, there were a couple of items that piqued my interest last week. The first came from GRIDtoday favorite John Powers (Digipede CEO), who couldn’t help but let slip a few details of the upcoming Digipede Network v.2.0. I have yet to speak with someone who was anything but supremely impressed with 1.x versions of Digipede’s product, so if the improvements to which John alludes are as great as promised, I expect uptake of 2.0 will be even greater – and I expect to hear from, if it’s possible, even happier customers.

The other blog that caught my eye comes from someone a little less-known (and by “less-known” I mean not known at all) in the grid community, software engineer Corey Goldberg. Corey pointed me to an overview of Google’s distributed infrastructure on http://highscalability.com, a site dedicated to the building of Web-scale architectures. What Google and its Web cohorts are doing with distributed technologies never ceases to amaze, although it probably shouldn’t if I were to stop and consider what it takes to offer a seemingly endless list of services while serving billions upon billions of requests every day. Anyhow, when the site is up and running (which it wasn’t when I wrote this), you can see Google’s specs here: http://highscalability.com/google-architecture.

In other news, be sure to look at this week’s entire table of contents, as we have a good number of big announcements from across the distributed computing spectrum, including ones from Cisco, Sun Microsystems and Univa, TIBCO, Oracle, CANARIE, VMware and 3Tera. The Cisco news, especially, is worth reading if you haven’t already, as the networking giant is looking to make waves in the virtualization marketplace.

Finally, as I noted last week, be sure to tune back in over the next couple of weeks as we present special coverage around the Next Generation Data Center event taking place Aug. 7-9 in San Francisco. We’ll have some pre-show features next week, and I’ll be sharing my thoughts about and experiences at the conference in the following weeks.


Comments about GRIDtoday are welcomed and encouraged. Write to me, Derrick Harris, at editor@gridtoday.com. 

Share This