I read Robert's post yesterday on the way companies are changing the way they measure performance from the current per-MHz/GHz to performance per watt. Robert's comments were based on the front page story yesterday, Power struggle: How IT managers cope with the data center power demands.

The result got me thinking about how we measure performance and what the yardstick should be.

My first comment is that power has long been a problem for data center managers, and I'm not going to repeat my own feelings and findings on the use (and often waste) of power, or the eco-friendly options that are available. I won't make the comment either about Sun's T1 CPU (Niagara) which was designed with high performance and low power requirements in mind and Sun's aptly named the servers CoolThreads. OK, I just made the comment anyway, but it nicely leads me on to part of the point I want to make.

I read Robert's post yesterday on the way companies are changing the way they measure performance from the current per-MHz/GHz to performance per watt. Robert's comments were based on the front page story yesterday, Power struggle: How IT managers cope with the data center power demands.

The result got me thinking about how we measure performance and what the yardstick should be.

My first comment is that power has long been a problem for data center managers, and I'm not going to repeat my own feelings and findings on the use (and often waste) of power, or the eco-friendly options that are available. I won't make the comment either about Sun's T1 CPU (Niagara) which was designed with high performance and low power requirements in mind and Sun's aptly named the servers CoolThreads. OK, I just made the comment anyway, but it nicely leads me on to part of the point I want to make.

As CPUs move to multi-core architecture, companies - both OEMs and end users - are facing issues of licensing. If you have a multi-CPU machine, you often have to buy a multi-CPU licence.

The query is whether a multi-core CPU is really two (or more) CPUs, or just one?

Microsoft have already made the decision, classing a CPU as a CPU, no matter how many cores it has. The same model is being used by VMware for their software.

Today marks the last day of my current job, because officially tomorrow, or more practically, Monday, I start a new job.

My current job is freelance writer and consultant, a job I've happily held for over 6 years. I work, almost, entirely from home, except for those times when I visit a client, and that freedom means that I sometimes start work as early as 3am, and have free reign over the jobs I do, when I work, what I write about and what hardware and technology I use for my own needs and that which I suggest, recommend and support to clients.

My new job will be Technical Writer for MySQL, and although my job and main focus of work will change, other aspects of my work will not. I still, for example, get to work from home. I'll be sitting at the same desk, initially using the same computers and still, on the whole, writing documentation. Most of my fellow workers around the world - for MySQL is very much a global company - will be doing the same. At our recent conference in Sorrento we had people from as distant as Australia, Seattle and Canada and as a near as Russia, Bulgaria, UK and of course Sweden, the home of MySQL.

Apple has released an update for their recent iPods that enables you to limit the maximum volume output, including the ability to lock it with a combination code.

Will this stop the problem?

In a few cases, maybe, but as I've argued before, this isn't going to make any difference to those people who want to play their iPod at maximum volume, and unfortunately that is the same group of people who will never install this update, and never set the maximum volume even if they did.

Apple has released an update for their recent iPods that enables you to limit the maximum volume output, including the ability to lock it with a combination code.

Will this stop the problem?

In a few cases, maybe, but as I've argued before, this isn't going to make any difference to those people who want to play their iPod at maximum volume, and unfortunately that is the same group of people who will never install this update, and never set the maximum volume even if they did.

I first picked up on it here, but the news is that Microsoft are delaying Virtual Server 2005 R2 SP1 until 2007 - with no firm date in that year.

As the Ars Technica piece mentions, Microsoft have already tried to grow market share by lowering the cost of their products, while VMware have just released their two key products - VMware player and VMware server for free.

Virtual Server is still my favourite solution for virtualization because of the security aspects, and the new release has some useful features, like live backups, as part of the feature set. I'll happily wait, but the delay makes me uneasy.

I've resisted for a while, but this post inspired me to comment.

Vista is a product that has been some time coming. In fact I can remember having discussions at Microsoft back in 2003 about the product.

Microsoft have trimmed features - check this item back in 2004, back when Vista was better known as Longhorn. With that piece, Microsoft trimmed those features to make a 'First half 2006' release date.

Now we hear its been pushed back, effectively by another year.

As if my earlier post on printing costs didn't convince you enough, there's a great fragment in today's Shark Tank on a simple waste of paper that could have been resolved with a little bit of user education.

I was reading an article in Information Age this weekend on Printing Costs. The article is about the costs and management solutions available for sorting out the costs of printing.

The article starts out with one of those scary statistics:

Each day in the UK, 523 million pieces of printed paper are thrown away by office workers.

I'll ignore the 'thrown away' comment - I really hope that a large quantity of that is recycled, but whatever the destination, the figure is still scary. 523 million sheets of printed paper are thrown away each day. Whether that is mistakes, reports that needed edits, or any one of a million other possibilities doesn't matter - that is a lot of paper, a lot of toner or ink, and vast quantities of printer wear and tear that is being wasted each day.