transcript
A meme saying:
"STOP DOING COMPUTER SCIENCE"
- Computers were supposed to solve math, NOT to be programmed
- C is a LETTER, not a language
- Wanna print() something? Write it in a PAPER with a PEN
- "I'm writing a recursive method with threads to optimize the CPU usage in a 0.02%" THIS IS A NONSENSICAL STATEMENT MADE BY DERANGED PEOPLE
Look at what PROGRAMMERs have been demanding your respect for, after all the led lights we put in their computers: (This is real COMPUTER SCIENCE, done by real COMPUTER SCIENTISTS)
- FUNNY COLORED LETTERS (with a picture of syntax-highlighted code)
- 178 COMPILATION ERRORS????? (with a picture of compilation errors
- A FAKE TEAPOT YOU CAN'T USE (with a picture of a 3d rendered teapot.
IF PROGRAMMING WAS REAL HOW COME NOBODY THOUGH IN DOING
while(true{print(money);}
end transcript.
Developer here: can confirm. (#4 is actually true.)
Indeed, anytime I optimize code, it ends up taking more CPU to run.
Isn't that what you're optimising? Trying to reduce runtime by increasing CPU load
If that's a genuine non-sarcastic question that isn't whooshing me then no - there can be other things like memory/disk usage... but if I'm optimizing for CPU I want it to use less overall cycles. It may be that the easiest fix is to throw money at the problem (always a fair option) which would mean getting a beefier/more processor cores to make the performance acceptable but this would usually just shift how cycles are being used to process them faster.
My joke above was that it'd use more total cycles which actually is generally the case if you're solving a problem by throwing resources at it (since you're likely incurring more overhead) but generally when you optimize you want to reduce the total number of cycles by somehow locating and eliminating work that doesn't need to be done.
It was genuine, in my workplace the only optimisation is for run time
I once made a multithreaded code for an embedded device.
Tested it on my development Desktop PC and it used all 4 cores.
Running it on the target device, all threads ran on the same core because it was too powerful for Linux to feel the need to use multiple cores.