sensible, sustainable business matters

Google's strategy is clear enough, and it's a good one. They have fingers in all the pies, and a winner in Orkut. But they have a problem. Orkut is big in the wrong places, and not big enough in the markets that to Google (and its shareholders) really matter. It's "first-world-centric" and wrong, but money talks, yes?

So here we are, mired in a world of seemingly endless choice, with social networks all around us battling for the right market to share and the "un-social" networks desperately adding social-like features to boot. Given the popularity and the potential it's inevitable that an aggregator would come along to join the pieces and gang up on facebook. And who better than Google? After all, if you belong to 2 or 3 social networks already, as many of us do, we feel the pain of having to log into different, 'fenced-off' systems. Whilst OpenSocial has some big buddies on board already it still has to woo developers, as well as beat facebook and its locked-down, non-transportable markup. So it's not going to be a quick kill.

Indeed it will be an interesting tussle. But I'd bet on the most integrated approach winning.
And that's Google's plan, too.

Labels: facebook, google, orkut, social networks

One impact of global climate change will be the need to conserve resources - and one way to do that is by consolidation. In some ways this will seem like going 'back to the past'. For example more consolidated travel by public transport, less of the individual transport we have become accustomed to. The IT field is awash with individual manufactured items - from PCs to MP3 players - and unless the power needs of these devices can be met efficiently (perhaps more solar panels on MP3s and mobiles) then they too will need to be re-thought. Of course just making them is a resource hungry process, too.

The Australian Financial Review reports on IBM's mainframe resurgence. I work for IBM and own some shares, but this is not necessarily my opinion, or that of IBM itself. In any event, here's a bit of the article: Mainframes return to the main game, 30 April 2007, by Joshua Gliddon. "Some 130 kilometres east of Portland, Oregon, in a small US town called The Dalles, search engine company Google has built a giant data centre on the banks of the Columbia Rover. Its reasons for doing so have less to do with cheap labour and land than with access to cheap and relatively clean hydro-electric power.Google is not the only company building data centres where power is cheap. Both Yahoo! and Microsoft are in on the act. The latter has committed $US2 billion ($2.4 billion) to its data centre strategy. Many data centres are made up of thousands upon thousands of cheap, replaceable commodity computers. These computers can chew through huge amounts of power and generate extraordinary amounts of heat, The cooling towers in Google’s The Dalles facility are four storeys high."

Which highlights one problem - massive numbers of individual servers out there, all drawing power. And IBM's solution? IBM's Bill Zeitler said that, "instead of putting together lots of servers in a data centre, companies would move to high-end machines, or even mainframes that could “virtualise” or pretend they are lots of computers running on one box." And he went onto say, “Clients are telling us that they simply can’t get enough power into their data centres to deal with the increasing number of servers,” Mr Zeitler said. “If you can do 60 times the work using the same amount of power then there’s an incentive to move to the high-end UNIX or mainframe environment.”

Let's see what Google does to solve this problem. They are famous for re-using and chaining together massive numbers of 'el cheapo' servers in a fault-tolerant network, rather than designing or simply buying a high-end machine or 2 to do the job. It's a strategy that has paid dividends for Google. They may have more ideas up their sleeves... or they too will be looking at consolidation and vitualisation.

Labels: google, IBM

As a search engine, PC World reckon Google wins - just. They also suggest that their search-engine competitors are doing some innovative things that go beyond mere lists of results on a page. Check it out here.

Labels: google, links