It's really not all that surprising how often people get confused about processor performance since unless you've followed the evolution of processors for the past few years or have really done your homework it can be quite confusing. Chip architecture means everything, and raw Mhz simply isn't an accurate assessment any more. Pentium 4's had very long data paths. It took more processor cycles for a single instruction to be executed, and thus they relied upon their faster clock cycles. Intel has since turned this around by revisiting the Pentium III architecture and heavily modifying it to produce the Pentium M's and the subsequent Core Duo's and Core 2 Duo's. They cut the processor cycle speed for the new chips but gained in performance. There's no easy way to compare the processors either, since some instruction types saw a greater boost in speed than others, and you'd therefore see a larger or smaller increase in performance depending on the application. As an example, the Pentium M's and Core Duo's have a greatly improved floating point unit compared to the Pentium 4's. An application using lots of floating point calculations would see a greater benefit in performance between the architectures than one that mostly relied upon integer calculations.
Back in the days before the Pentium 4, AMD and Intel CPUs were fairly comparable when looking at clock speed. An 800 Mhz Pentium 3 and an 800 Mhz AMD Athlon would have similar performance. Once the Pentium 4's came out and were clocked at much higher speeds, AMD came up with their new model number scheme for their Athlon XP processors which rated them at the approximate speed equivalent of a Pentium 4. Thus an Athlon XP 2400+ was supposed to have the equivalent computational ability of a Pentium 4 2.4 Ghz. This was done since many consumers were under the impression that a higher speed in Mhz meant more processing speed and power. So the numbering scheme would hopefully help a consumer compare processors based on performance rather than clock cycles per second. Whether this helped or caused more confusion is debatable, and certainly their numbers didn't always match up with benchmark comparisons.
It's because of these differences in architecture that it's impossible to really estimate how well a particular processor will perform in a given application without specifically benchmarking the system using the actual application. Thus without running NWN2 and benchmarking various systems we can't do much more than make an educated guess at how a system will perform. I've run the app at that systemrequirementslab.com address on both my notebook and desktop computers and they've come out about how I expected. I would say it's a fairly accurate assessment given the info we've received so far from developers. But there's certainly some room for error. Honestly, that link should be stickied somewhere in the NWN2 forums. That way people like the one quoted previously with a Core Duo 2.16 Ghz would realize his processor not only meets the minimum requirements, but the recommended as well, even though it's operating at a lower frequency.
It'd be nice if there were some easy way to eliminate this confusion. Maybe developers should create a web app that simply says 'run this to see if your system will run the game'. Since listing out system requirements clearly enough to deal will all the different architectures would take up most of the space on the box.
Oh well, enough of my rambling. If you see people asking whether their system will run NWN2, just point them to that link. It'll probably save a lot of time and confusion.
