Security: reward the simulator detector, punish the real thing.
2 for the price of one. Details in the blog references:
Product A detects non-viruses as viruses better than product B, therefore Product A is better!
Very interesting. Thanks.
The entire IT industry is based upon the premise of inappropriate testing.
Hardware review(/supplier-bribe-recieving) sites especially excel at
this If there's one thing that is a valid measure of how fast a PC will
be in day-to-day use, it's calculating Pi to X million digits. We all
do that all the time, right? I for one do nothing else. Even when they
claim to base benchmarks on real software (naming no names), and it
just loops through the most P4-optimised photoshop filter 500 times
(all us artists just sit repeatedly applying Gaussian blurs, right?)...
We all recognise the merits of testing a graphics card using an engine
(3DMark200X) that's not used in ANY actual games, right? I didn't buy
that Radeon to actually play games or do 3D work... no, I purchased it
to loop through pointless demos faster than the other guy.
Inappropriate and pointless testing isn't just insane, it's also the industry standard
Well, that's why pages like Tom's Hardware really took off, despite all they really did was run real-life scenarios. They actually benchmarked the games, and took real live tests. Nothing special besides that.
I'm a little disapointed with their latest tests on CPUs though, it seems like they used machines that are way more expensive than most people would want to afford. Still, the tests are still usefull, as far as the CPUs.
I really like their benchmarking the power consumption - the game will likely run just fine on the either CPUs but you will notice your hydro bill if you run the machines all the time ;P