Figure 4-17 shows how we model both a CPU subsystem and an IO subsystem. Even a quick glance shows they are fundamentally different. The key in understanding their differences is recognizing that an IO device cannot serve every IO request. In fact, an IO device can service only requests that are intended specifically for that device. So while one IO device may be actively servicing requests, it's completely possible another device may not be servicing any requests. In stark contrast, a CPU core has no choice (not that this represents a problem either) but to service the next transaction in the single CPU run queue.

Figure 4-17. CPU and IO subsystems are fundamentally different. In a CPU subsystem, any core can service any transaction. An IO subsystem device can service only requests directed to itself. This enables hot and cold IO devices. Even with advanced software, the balancing activity cannot be perfect.

This natural seemingly imbalance of IO activity is why so much effort and expense is put into keeping all devices equally busy. But even when all devices on average are just as busy as the next, at a point in time, they can fall into the situation shown in Figure 4-17. This is easy to see with an animated graphical simulation and much more difficult to describe in words, but I'll try.