3 ms·
I totally disagree, I think that era has been over for the last 20 years. I keep coming back to a talk by Daniel J. Bernstein, that the assumptions computer sc
by Gibbon1 8y ago
I totally disagree, I think that era has been over for the last 20 years.
I keep coming back to a talk by Daniel J. Bernstein, that the assumptions computer science teaches about performance are wrong. His claim is for real programs most code is ice cold and a tiny percentage is hot. And that's getting worse not better as time goes on.
What important is; correctness, latency, compile time. Not raw speed.
- adrianN 8y agoIn my experience that claim is very true before you fire up your profiler. But once you've optimized your program for a bit the profile gets a lot flatter.
- Gibbon1 8y agoI think if your programs profile is flat it means none of your code is actually in the hot path. It's buried behind library functions and OS calls.