Efficiency still matters

Writing efficient code is, arguably, the problem which spawned the entire field of computer science. It's a hard problem, obviously.

However, before you can start working on this problem, you must first solve the problem of writing code that works. Since one of the virtues of programmers is laziness, programmers sometimes fall into the trap of deciding that once the problem of writing correct code is solved, the programmer's work is done. Frequently Moore's Law is invoked as an excuse: "The processor speeds are so fast that the wasted time is negligible."

Part of this mindset can be explained when you realize that most programmers were trained on small problems, where this rationalization is in fact correct. There are many problems that the Ancient Masters optimized to death, and that student programmers must study, that can now be safely solved using brute force.

Practically all of simulation lies outside this domain.