I work in HFT and all I ever talk about is cache hotness and hardware marshaling. I haven't given a shit about a loop or AVX in like 5 years. "Performance" is pretty disparate and application specific.
-
-
-
Similar story in scientific programming. Effective CPU usage matters a lot, but saving I/O latency by doing DMA transfers direct from disk to Ethernet card without passing through RAM often matters more.
End of conversation
New conversation -
-
-
Humble performance.
-
It's the truth, unfortunately. I've worked with enough good performance programmer to know I am not a good performance programmer :(
- Show replies
New conversation -
-
-
Who is first?
@rygorous ?Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I would really love, if some of the RAD people (
@rygorous, please :D) would make some videos, etc. to show us their wizardry That skill needs to be taught to the public -
Fabian already has some excellent videos you can watch on modern CPU architecture: [1/4]https://www.youtube.com/watch?v=oDrorJar0kM …
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.