Every industry measures performance, we just use different jargon to discuss it. When we talk about the raw power of a car engine we talk in terms of horsepower; of harnessing the performance of hundreds of horses such that they work together as a single unit. In the world of computing we use terms like MIPS (million instructions per second), and in the world of application delivery we measure performance in terms of transactions per second (TPS).

The problem in the world of computing is, unfortunately, that simply adding more "horses" to the mix doesn't linearly increase performance. Generally speaking, adding another CPU or blade to a system doesn't result in a nice, linear increase in performance due to the underlying management and software necessary to harness the power of that additional component. The rule of diminishing returns tells us it's just not possible in the computing realm.

Ahhhh, but rules were meant to be broken.

Imagine a system that actually does linearly increase performance with each new CPU that's added. Imagine a system that, like a car, harnesses the full potential of all processors and treats them as a single system rather than as individually managed entities.

Imagine a system built not on traditional multi-processor management techniques like SMP (Symmetric Multi-Processing) that yield diminishing returns with each additional processor added to the system, but a system that virtualizes each and every processor in such a way that the performance of the system increases linearly with the addition of each processor.

Imagine the performance you could get out of a system like that. Imagine the traffic you could manage, the applications you could serve, and the video you could stream. Imagine millions of transactions per second - all from a single, unified delivery controller.

Now imagine that you can add this kind of power on demand. Without reconfiguration, without rearchitecting your network, without adding yet another device to your infrastructure. Imagine the ability to scale your infrastructure without introducing the complexity that comes with traditional scaling techniques. Imagine deploying not two or three tiers of infrastructure and the tangled web of cabling and routing that goes with it, but just one. A very powerful one that scales effortlessly. As though it was designed to do just that...

Instead of being bound by old rules, imagine breaking them down and setting new ones. Imagine performance as you've never imagined it possible before.

Stay tuned ... imagination is about to become reality.